self-control, inhibition, & neuroscience and volition & free will response
Camille Barnes
The Brass and Haggard and the Walsh, Kuhn et al articles both looked
at similar processes, but one used fMRI and the other used EEG to
explore these processes. The initial study (brass and haggard)
located the dorsal frontomedial cortex and left and right anterior
ventral insula and the right superior temporal sulcus as being active
in inhibition of action, and the primary sensorimotor cortex and the
cerebellum as being active when intended actions occurred. The Walsh,
et al article focused on the timing of when these different
activations occur.
In both the Brass and Haggard articles, I am confused about their
experimental procedure. So I think how it worked was before the
studies, participants were told you are going to press a key, but
first you have to look on a clock and decide when you intend to press
the key and also, on some trials at the last second you need to
inhibit the urge to press a key. Putting myself in the participants
shoes, (a) because of lay theories that intentions come before
actions, I would always report that my intention came several seconds
before the action, regardless of whether or not this is when I
actually first intended to press the key (but in actuality it seems
the intention would occur as soon as they hear instructions that they
are supposed to press a key, right?). (b) it also seems that for
trials in which you inhibit the action right before you do it, you
would have to go into that trial thinking “ok I haven’t inhibited in a
while and I am supposed to inhibit about half of the time, so I will
inhibit this time,” which obviously does not display a last second
inhibition. (c) If people really are deciding to inhibit at the last
second (which I doubt is really occurring) then there should be an
error rate, in which they intend to inhibit but fail and still press
the key. Similar to this it was not stated how they determined
inhibition trials, did they use self-report of trials that
participants intended to inhibit (this would include trials in which
they were unsuccessful at inhibiting) or trials when no action
occurred (which would not include any error). I just wish they would
break down this procedure and show how each step is really showing the
process that they say it shows. These studies do show differential
brain activations, so it is obvious that different processes are
occurring, I am just not sure what these different processes are and
am not convinced that they show intention to act and intention to
inhibit action, as the authors suggest. I liked how they are using
different brain scan technology, with different pros and cons, to
examine the same procedure, this show a clearer picture of the timing
of brain activations and a more precise look at the brain activations
themselves.
Inzlicht and Gutsell looked at stroop performance following either
emotional suppression or no emotional suppression. They found that
those who suppressed emotion performed worse on the stroop and this
was mediated by decreased activity in the ACC. I liked how this study
was simple and had used well established methods (emotion suppression
during video and stroop) as self-control measures. Also, I liked how
they had a high number of participants (more than previous brain
scanning studies that we have seen. One interesting thing was that
when viewing the bar graphs, it appears that there was little
difference between groups on number of errors during stroop task, but
there was a much larger difference in response latencies between
groups. So it seems the ACC is still able to monitor effectively for
errors even when it is fatigued from previous self control tasks, but
that it just takes longer. Overall I like how this study, took a
procedure similar to that of a lot of self-control studies and added
the additional piece of evidence, brain activation, to the theory.
This type of study strengthens the argument that the theory expresses
real processes.
I liked how Soon et al analyzed their data. They didn’t just look to
see what was active, but they looked at how predictive the activated
of each brain area was towards the corresponding action. This helped
them to really parse apart the different steps and what brain parts
are involved with each step. They found that the outcome of the motor
decision during execution of the decision was predicted by the primary
motor cortex and the SMA. The frontopolar cortex and a portion of the
parietal lobe were predictive of encoding the motor decision ahead of
time. The SMA and pre-SMA seem to be active long before a decision is
made (about 5 s), while the frontopolar and parietal cortex are active
just before a decision is made.
Stuart Daman
don't feel many of the authors we read this week do a good job of
saying what they mean very well or clearly. For example, what the heck
does this mean: "it remains unclear whether the endogenous decision to
execute or inhibit an intended action involves an analogous system for
endogenous inhibition of intentional action." (Brass & Haggard, p.
9141). Or how about "The freely paced button presses occurred, on
average, 21.6 s after trial onset, thus leaving sufficient time to
estimate any potential buildup of a 'cortical decision' without
contamination by previous trials." (Soon et al, 2008, p. 543). These
people don't write good. I think quite a bit of it may have to do with
the 'short-report' nature of the places they got their studies
published, but this says something itself. We might wonder if their
work wasn't good enough to be published in a full length article in a
better journal. I don't wanna bash Science magazine, but I thought it
was intended for a wide variety of audiences, and these articles are
definitely not. At any rate, in order to communicate clearly much of
this information, more space is needed to clarify things like
intention and the temporal organization of this stuff, and these short
reports do not do it justice. On top of that, reading and making sense
of four or more of them in one week is tough.
In this vein, I found several of this weeks' articles full of mumbo
jumbo. In the Lau et al (2004) paper, the first few paragraphs flowed
well and the final one summarized the paper well, but the rest of it
was a mess, as far as I could tell.
Any of the papers with Brass or Haggard as an author seemed doomed and
plagued with bad phrasing and terminology.
Another thing about the small sample sizes of many of these studies
occurred to me while reading Brass & Haggard (2007) too. In many
papers, the number of participants are weird numbers. For example, in
this paper, they gathered data from 19 participants, but only used
data from 14 because the other 4 didn't follow the instructions and
one for moving their head. I can't help but wonder if researchers in
these studies gather data from as many participants as necessary to
find the effects they are looking for. It is, perhaps, a bit more
understandable because of how 'expensive' it is to run participants in
some of these studies. I would be more convinced if every study had 16
participants, or better yet 20. But that there is variability and the
numbers are small, it seems like they stop gathering data as soon as
they find what they are looking for. I suppose I should be clear that
we don't know that this is the case, but it is a suspicion of mine
that cannot easily be abandoned.
I am a little torn about how I feel about the Brass & Haggard (2007)
study, in more particular terms than mentioned above. How do they
define intent? The participants in this study were to develop an
intention of pressing a button and then stop it. What if they decided
whether or not they would press the button during the ITI, when was
the intent developed? Or what if they decided it in the beginning of
the trial? OR is intent the sending of a motor signal? In this last
case, ceasing the intent would mean stopping the movement after it has
already began. Or is intent something between these? I also don't
think it's OK to leave the defining of intent up to the participants,
because I don't see how it can't be subjective. Camille talked about
this too. However, I think they were able to distinguish action and
intention trials by whether or not the button was pressed (yes in
action, no in intention). I also couldn't tell if their stuff was
nonindependent or not. I feel like there is usually enough information
to guess with some confidence, but I couldn't really tell at all in
this one. And who the hell says "median" when referring to 'medial'
areas? Median is the central score in a set, not a region of the
brain.
I took a special topics course in my Master's program called "Mind,
Body, Action", which focused almost entirely on motor imagery. In that
course, I read the Blakemore & Decety (2001) article that was included
for this week. In particular, that article discusses the "forward
model" which I had loosely floating in my mind during one of our
discussions a few weeks ago in class. I think it's a fascinating
concept that our minds use previous information to predict what the
outcomes will be of our motor signals, while including 'on-line'
sensory information and using it to modify the current motor signals.
One example that may help illustrate the idea is when you try to pick
something up, drop it, but manage to catch it during its fall. The
original motor signal moves your hand to the object and grasps it,
using previous experience to gauge the amount of movement necessary to
reach the object and get a sufficient grip on it. Say the object is a
bit heavier than expected or you are not attending to it 100%, thus
dropping it because your hand should have gone one inch to left or
because of insufficient grip. Your sensory neurons recognize this--the
object is not clearly in hand as expected--and can feel it missing or
leaving your grasp. This information, in turn, is used to amend the
motor signal in such a way to get a complete grasp on that object,
with minimal delay. Actually mapping this temporally can be difficult
because the motor and sensory neurons are working simultaneously, in
parallel, and are probably communicating to one another (within or via
the brain) the whole time.
The self-control failure paper (Inzlicht & Gutsell, 2007) was
interesting and much more cohesive than many of the others. One
shortcoming I noticed was that the authors explain that the ERN is
indicative of the conflict-monitoring system, and provide evidence
supporting this notion. However, they provide no clear reason to not
suspect that it is also indicative of the regulatory system. How do we
know that the ERN does not indicate both conflict-monitoring and
regulation? How do we parse these systems/processes in brain activity
terms? Their manipulation checks are also unfortunately crap. When
instructed to to do something, of course people are going to report
having done it. Their questions do not implicitly measure whether or
not their participants actually did so. I was also confused at the
fact that there was no main effect in their behavioral measures on
suppression/control. The graph makes it look like there should have
been, and many other studies would predict it too. Overall, I thought
they make a very interesting and valid point about the course of self-
control (failure) in the brain; but as with much other neuroscience
research, it asks more questions than it answers.
I also realized later that I had already read the Amodio et al (2008)
paper. Taking a commonly employed model of motivation and showing that
it has reliable neural correlates is a big step forward and paves the
way for oodles of studies and research questions. I originally read
this article when I was concerned with drawing similarities between
that dual-system model of motivation (i.e. BIS/BAS or approach/
avoidance) and self-control strength. I'm not sure whether or not it
was intentional, but this article makes many comments that allude to
self-control in the way we commonly conceptualize it in our program.
beka strock
SOC NEURO WK7
The best angle to view this week’s topic I thoughts, was presented in Haggard (2009)* in its emphasis of the frontal coretex as the seat of self-regulation, although it also commented on newer directions of research in its discussion of the role of the parietal lobe as a predictor of future consequence. One question raised in this article was why lessioned parietal lobe patients would increase volition (rather than experiencing the predicted lack of volition)—what if they over estimate the volition of observed actions because all actions feel nonvolitional but are attributed to being volitional normally, thus they would estimate any action (to include external) as feeling the same as their own action. Then the lack of volitional feeling would be accounted for since the patients would still exhibit a lack of difference in response to what would normally be attributed to two different types of actions (volitional/nonvolitional).
The BIS/BAS systems really did not clarify the understanding of self-regulation processes for me at all, the only point I got out of it was: BIS is bottom up/monitoring and is primarily located in the ACC whereas BAS is top-down…and located where, (frontal somewhere, but with asymmetrical activations)? I’m not sure if the terminology of BIS/BAS instead of common English was throwing me off, but I found this article (Amodio et al, 2007) entirely unintelligible.
My main issue in the remainder of the articles was trying to keep straight who allowed for free-choice versus semi-free choice etc., since almost each set of authors complained that it was rarely addressed in the research. A lot of these seemed to address sub-processes that might contribute to free-will but I thought their claims to have found the neural substrates of free will in the introductions/ conclusions, abstracts and titles were far to broad.
Soon (2008) allowed choice of button, but not whether or not to press. Main takeaway here is the emphasis on top-down processes of decisions (i.e. chronological origination in frontopolar coretex first).
Lau et al (2004)* allowed for timing of finger movements based on clock times and examined activation related to attention to intention (pre-SMA, R-DPFC, L-IPS).
Brass & Haggard (2007)* inhibit response (key press) which they decided the time to execute before each trial, but inhibited based on a tone (not free choice)—I think. Not sure if I understood that correctly. (Sidebar to this theme: The explanation of why of why inhibition is top-down here was pretty on pretty shaky ground to me, the reasoning being that “more anterior regions are related to intentional actions.” Great justification, there buddy…or, not so much. The authors did cite another publication, but still for that sort of claim, seriously throw us a bone!) Regardless, activations attributed to inhibition were found at fronto-median cortical areas DISTINCT from CMA/pre-CMA areas attributed in other studies…which I guess they just decided to call dFMC. Just label the place, and it exists, right? Holy crap. (Sorry I am feeling quite sarcastic this week!)
Sidebar: Nice to see something other than FMRI! Go EEGs! J
Walsh et al (2009) I liked the choice-paradigm the best since participants actually had full choice latitude in deciding when to respond, when not to respond, as well as when to prepare and inhibit responses. I was slightly confused as to how/when their reporting of when they intended to act or inhibited, versus the actual act, could be compared since the recall measures have got to be lacking in temporal accuracy in comparison to measured reaction time. Relevant self-control activations: ACC used for actively inhibiting but not for non-inhibiting (action). Despite its applicability to self-control, this article seemed much more focused on the decision-making process than the control mechanisms.
Inslict & Gitsell (2007) (brownie points!!) used the depletion model to study activations! YAY! I also liked that they split the processes into monitoring and operating, but I would say that this can only be one part of the explanation, what about the active over-riding-response self-regulatory (“operating”) strength? I don’t think this evidence rules out that the operating process is also likely moderating this depletion effect as well rather than only monitoring. As expected, these responses (or lack thereof) were mediated by ACC EEG activity. Finally, its cool that there is have a specific signal (ERN) that only arises from this particular system (dACC), although I would be interested to have more information to understand why that signal could not be originating elsewhere in the brain.
Lindsay Morton
For Brass and Haggard (2007), let’s start with the good points. First, I like that the authors situate their work in a human-centered/societal context from the start. I also think the inclusion of the control-tone blocks was important for their demonstration that participants prepared intentional actions (and thus activated the pre-SMA) in both inhibition and action trials as compared to tone trials. The research also points to individual differences in inhibitory tendencies at the neural level (as shown by results on page 9143 and clinical discussion point on page 9144). At the same time, I have to agree with some of the critiques that Camille posted and I also question the external validity of their inhibition paradigm. More specifically, Brass and Haggard (2007) asked participants to endogenously inhibit responses, and I’m not convinced that this can truly occur. In an evolutionary context, the ability to inhibit one’s urges grew from the need to belong to a social group. Thus, it seems likely that most natural inhibitory behaviors arise from adherence to social rules and regulations. Do you think there is ever a situation in which inhibition is not externally cued or socially influenced? Even in this experiment, the participants may have been prompted by signals in the environment and their perceptions about experimenter wishes.
The Soon, Brass, Heinze, and Haynes (2008) piece seems to be a great one for discussions of fate versus free-will, and although this is more of a social-cultural topic, rather than a scientific, psychological one, it’s still interesting. Basically, the researchers explained that before participants are aware of their own decision, their brains have already been unconsciously activated for up to 10 seconds. They posit a fairly exact model too, which I just wish they tested with mediation analysis or SEM. My only other critique is that the experiment seems to require that participants keep a lot of information in mind (i.e., button to press, finger to use, urges, screen choices, etc), and thus the activation seen by the researchers may be a map of not only of the conscious decision (as the researchers thought) but also intention, action preparation, and other memory processes.
While reading Lau, Rogers, Haggard, and Passingham (2004), the authors explain that the greater pre-SMA activation seen in the intention condition participants, as compared to the movement condition participants, could be due to the greater difficulty of making judgments about the timing of intention. It was unclear to me how they refuted this idea (i.e., “inspection of the individual adjusted raw data for time course of hemodynamic responses at peak coordination of the activation suggests that this is not the explanation for the result”). Maybe someone in class could better explain, but it seems that differences in task difficulty still remain an alternative explanation for the differential activation. Lau et al. (2004) also suggest that one needs to attend to intentions for conscious control of action. At the same time, would it be possible to be unaware of intentions yet still control action? Specifically, work with non-conscious goal priming has shown to have direct effects on goal-directed behavior (e.g., Bargh et al., 2001). I would consider all goal-directed behavior to be consciously controlled, yet according to Lau et al. (2004), would it be possible to act on intention without attention to it?
Inzlicht and Gutsell (2007) addressed the hypothesis that self-control depletion impairs the monitoring/error-detection system. They found that engaging in a self-control task impairs subsequent self-control task performance by weakening the neurally-based monitoring system, as evidenced by weaker ERNs in depleted participants. In other words, compared to control participants, depleted participants were less able to monitor/detect their own losses of control (i.e., “less responsive to a mismatch between their actions and their goals”). In contrast to these results, Schmeichel (2007) has shown that memory maintenance operations, which are conceptualized as keeping goals or other information active in working memory, are unaffected by self-control depletion. In a recent study examining how self-control depletion affects rational and experiential/emotional processing, I found that depleted participants reported using memory maintenance strategies to a greater extent than control participants. This led me to start considering that one impact of self-control depletion might be either greater reliance on memory maintenance because of possible inability to engage in memory updating/rerouting OR decreased ability to inhibit irrelevant information as shown by greater attention to all information presented. This second possibility should be related to decreased anterior cingulate activation, which has been seen as involved at the level of inhibition of irrelevant information (e.g., Smith & Jonides, 1999). I’d be curious to know how you all think these results/ideas coincide or contradict the findings of Innzlicht and Gutsell (2007)?
David Dinwiddie
The Amodio et. al (2008) article looked at the neurocognitive differences in Behavior inhibition and behavior activation systems (BIS/BAS). The authors were also trying to clear up confusion on whether BIS was actually an inhibition response or whether it was instead engaging in avoidance behavior. The results showed that there was greater ACC activation in no-go or inhibitory trials. Activity in the ACC is associated with conflict and so the authors conclude that this means BIS is an inhibition of behavior rather than an avoidance behavior. Also a key conclusion, participants high on BAS had greater left-sided frontal cortical asymmetry, an area associated with an approach orientation. Therefore BAS is associated with a motivational system for coordinating approach/avoidance responses. This conclusion was a bit unclear to me. This was in the discussion section but it seems the authors may have gone a bit too far and relied on other peoples data too much in coming to this conclusion.
The article by Brass & Haggard (2007) states that there is a network in the human brain, located in the fronto-median cortex, which forms intentions about how and when to act. This information alone is very interesting to me. This seems like it may be the area which is always talking to you, telling you what you are thinking and what letters to type as it is doing for me right now. This area was activated equally when preparing to act in inhibition trials and action trials. dFMC activation was positivity correlated with inhibition. When there was a stronger activation of the dFMC participants were more likely to inhibit a response. The area that was associated with generating an intentional action is different than that area associated with inhibiting an intentional action. The dFMC is just the area for inhibiting this action not for generating an action. This seems to be a key area of the brain for inhibition and self-control. When this area is activated one is more likely to inhibit a response and the activation of this area differed greatly between subjects just as self-control can differ.
Walsh et. al. (2010) essentially reproduced the Brass & Haggard study with different equipment. (EEG instead of fMRI) Their results were similar as well. By using the EEG there were able to do a time-course check to make sure that participants were actually following the instructions of the experimenters not to make up their mind until the trial. One thing that was acknowledged in this article concerned me. They mentioned that 94% of the time, when given a decision to respond or not to respond, the participants chose the opposite behavior from the previous trial. This concerns me and makes me believe that the participants were nearly set in their decision from the time of the previous trail and just changed their mind on occasion. A number closer to 75% would make me feel better.
The Inzlicht & Gutsell (2007) article was very interesting to me and is very closely related to the work done in Mark’s lab. There were much smaller ERN’s for subjects that were in the emotion-suppression group than subjects in the control condition. This study provides support for the depletion model of self-control. The article did not do a good job in explaining the parts of the brain where the activation had actually decreased. They were in the conflict-monitoring/error detection system which is supported by the ACC and PFC. Perhaps the lack of detail in the locations is due to the study being an EEG study rather than an fMRI study. It would be nice to see the same study done again in this setting to see if even greater support could be given for the self-control theory.
matt gailliot
The topic of intention is difficult to communicate. Unfortunately, I
don't feel many of the authors we read this week do a good job of
saying what they mean very well or clearly. For example, what the heck
does this mean: "it remains unclear whether the endogenous decision to
execute or inhibit an intended action involves an analogous system for
endogenous inhibition of intentional action."
to stop reading the article? or to stop reading the email? that the article contains both true and false information? that it is not what it seems?
(Brass & Haggard, p.
9141). Or how about "The freely paced button presses occurred, on
average, 21.6 s after trial onset, thus leaving sufficient time to
estimate any potential buildup of a 'cortical decision' without
contamination by previous trials." (Soon et al, 2008, p. 543).
precedent is BS? if someone judges one based on perceived similarities to the past, a build up will occur? so, presumably the intent of the system is bad. the court system, in my experience, is very antagonistic. people don't talk considering the welfare of all. the defendant is a target. do they ask, 'what do you think?', 'do you think this is fair?'. some want the decision for themselves.
These
people don't write good. I think quite a bit of it may have to do with
the 'short-report' nature of the places they got their studies
published, but this says something itself. We might wonder if their
work wasn't good enough to be published in a full length article in a
better journal. I don't wanna bash Science magazine, but I thought it
was intended for a wide variety of audiences, and these articles are
definitely not.
too removed from the source? too removed from to whom the message is communicated? perhaps your domain is too different from the domain(s) for which it's written.
maybe they run until significance?
I am a little torn about how I feel about the Brass & Haggard (2007)
study, in more particular terms than mentioned above. How do they
define intent? The participants in this study were to develop an
intention of pressing a button and then stop it. What if they decided
whether or not they would press the button during the ITI, when was
the intent developed?
for bad intentions, if one acts in ways that produce a consistent effect, the response to it is of satisfaction and later attempts to produce the same effect, then wouldn't this be clear intent?
they do? i tried to convey to participants that reporting bad things was okay, they didnt have to try to make any particular impression on the experimenter(s) or through their responses.
matt gailliot
For Brass and Haggard (2007), let’s start with the good points. First, I like that the authors situate their work in a human-centered/societal context from the start. I also think the inclusion of the control-tone blocks was important for their demonstration that participants prepared intentional actions (and thus activated the pre-SMA) in both inhibition and action trials as compared to tone trials. The research also points to individual differences in inhibitory tendencies at the neural level (as shown by results on page 9143 and clinical discussion point on page 9144). At the same time, I have to agree with some of the critiques that Camille posted and I also question the external validity of their inhibition paradigm. More specifically, Brass and Haggard (2007) asked participants to endogenously inhibit responses, and I’m not convinced that this can truly occur. In an evolutionary context, the ability to inhibit one’s urges grew from the need to belong to a social group.
Thus, it seems likely that most natural inhibitory behaviors arise from adherence to social rules and regulations. Do you think there is ever a situation in which inhibition is not externally cued or socially influenced? Even in this experiment, the participants may have been prompted by signals in the environment and their perceptions about experimenter wishes.
The Soon, Brass, Heinze, and Haynes (2008) piece seems to be a great one for discussions of fate versus free-will, and although this is more of a social-cultural topic, rather than a scientific, psychological one, it’s still interesting. Basically, the researchers explained that before participants are aware of their own decision, their brains have already been unconsciously activated for up to 10 seconds. They posit a fairly exact model too, which I just wish they tested with mediation analysis or SEM. My only other critique is that the experiment seems to require that participants keep a lot of information in mind (i.e., button to press, finger to use, urges, screen choices, etc), and thus the activation seen by the researchers may be a map of not only of the conscious decision (as the researchers thought) but also intention, action preparation, and other memory processes.
While reading Lau, Rogers, Haggard, and Passingham (2004), the authors explain that the greater pre-SMA activation seen in the intention condition participants, as compared to the movement condition participants, could be due to the greater difficulty of making judgments about the timing of intention. It was unclear to me how they refuted this idea (i.e., “inspection of the individual adjusted raw data for time course of hemodynamic responses at peak coordination of the activation suggests that this is not the explanation for the result”). Maybe someone in class could better explain, but it seems that differences in task difficulty still remain an alternative explanation for the differential activation. Lau et al. (2004) also suggest that one needs to attend to intentions for conscious control of action. At the same time, would it be possible to be unaware of intentions yet still control action? Specifically, work with non-conscious goal priming has shown to have direct effects on goal-directed behavior (e.g., Bargh et al., 2001). I would consider all goal-directed behavior to be consciously controlled, yet according to Lau et al. (2004), would it be possible to act on intention without attention to it?
Inzlicht and Gutsell (2007) addressed the hypothesis that self-control depletion impairs the monitoring/error-detection system. They found that engaging in a self-control task impairs subsequent self-control task performance by weakening the neurally-based monitoring system, as evidenced by weaker ERNs in depleted participants. In other words, compared to control participants, depleted participants were less able to monitor/detect their own losses of control (i.e., “less responsive to a mismatch between their actions and their goals”). In contrast to these results, Schmeichel (2007) has shown that memory maintenance operations, which are conceptualized as keeping goals or other information active in working memory, are unaffected by self-control depletion. In a recent study examining how self-control depletion affects rational and experiential/emotional processing, I found that depleted participants reported using memory maintenance strategies to a greater extent than control participants.
This led me to start considering that one impact of self-control depletion might be either greater reliance on memory maintenance because of possible inability to engage in memory updating/rerouting
this also requires recognizing memory deficits, less likely with depletion? or effortful?
OR decreased ability to inhibit irrelevant information as shown by greater attention to all information presented.
is this an automatic process, in which case 'ability' might not be the ideal word. or perhaps it's strategy selection. . .