Monday, April 20, 2009

Neural Correlates of Admiration and Compassion and Envy and Schadenfreude

In light of all the sensationalistic press coverage about a journal article that wasn't publicly available last week, it's worth taking a moment to look at the actual experiment. Of course, the savvy skeptics know by now that the paper in question (Immordino-Yang et al., 2009) has absolutely nothing to do with Twitter (see Recommended Reading below for a recap). Instead, the authors conducted a neuroimaging study to examine the brain's response to stories designed to elicit the emotions of admiration and compassion. To do this, the participants (n=13) first watched a series of mini-documentary narratives about real people (who were not celebrities). Each of the 50 narratives was 60-90 sec long, and incorporated audio, video, and still images to convey stories categorized as:

1. Admiration for virtue (AV), which involved people performing highly virtuous, morally admirable acts. The narratives emphasized the virtuous and morally admirable nature of the protagonist, such as dedication to an important cause despite difficult obstacles, and did not include displays of notable skill.

2. Admiration for skill (AS), which involved people adeptly performing rare and difficult feats, e.g., an athletic or musical performance, with both physical and cognitive components. No physically or socially painful acts were shown, and the skillful feats, although amazing, did not imply a virtuous protagonist or reveal a virtuous act.

3. Compassion for social pain (CSP), which involved people in states of grief, despair, social rejection, or other difficult psychological circumstances. No physical pain was evident in these narratives, and the troubling circumstances were discerned from the descriptions, rather than being apparent in the images shown.

4. Compassion for physical pain (CPP), which involved people sustaining a physical injury. The injuries were caused by sports and other mishaps and had no moral or social implications. The injuries were not the result of malevolence, and the participants were reassured that the injuries had no long-term implications....

5. Control narratives, which involved comparable living, mentally competent people engaged in or discussing how they felt about typical activities under commonplace social circumstances. These circumstances were engaging but not emotion provoking.
After each of the narratives, the subjects were asked to discuss how they felt about the protagonist's situation. This part of the study took 2 hrs, and was conducted outside the scanner. For the fMRI portion of the protocol, 5 sec recaps of all 50 scenarios were presented, and the task was to:
induce in themselves for each story, as strongly as possible, a similar emotional state to the one they had experienced during the preparation session and to push a button to indicate the strength of the emotion they achieved in the scanner (from 1 to 4...). Participants were asked to report candidly on the strength of their current feelings in the scanner, rather than on the strength of feeling they remembered from the preparation session.
OK, so the subjects were first asked to remember how they felt 2 hrs ago, then try to duplicate that feeling, and then report on how they feel now (rather than before). So there's a memory component and a decision component (i.e., to not confuse past feelings with the present). Each trial was sorted post hoc on the strength of the reported emotion, and only the effective trials were included in the analysis.

The comparisons of interest were pain (compassion) vs. non-pain (admiration), and emotional responses to other peoples’ social/psychological conditions (AV, CSP) vs. to their physical conditions (AS, CPP). One of the first issues discussed is the recruitment of homeostatic mechanisms when experiencing these social emotions:
It is well known that basic emotions such as fear, sadness, and happiness and limited social emotions such as moral indignation engage neural systems concerned with sensing and regulating body function with varying patterns, and it has been hypothesized that among those systems, the insula plays an especially prominent role. It is also known that engagement of social emotions and the consequent feeling for another’s social/psychological situation are described by poets and lay people alike in visceral and bodily terms and in terms of their heightening effect on one’s own self-awareness or consciousness.
Basically, people may have visceral responses to the circumstances of others. How do these responses differ across physical vs. psychological situations? For example, admiring a gymnast's skill on the balance beam vs. admiring a student's charity work with Habitat for Humanity? Or feeling compassion for a single mother who loses her job vs. feeling compassion for one who sprains her ankle? Although it's not mentioned in the paper, this idea draws on Antonio Damasio's somatic marker hypothesis (e.g., Damasio, 1996). Perhaps this omission occurred because two of the main regions implicated -- the ventromedial prefrontal cortex (VMPFC) and the amygdala -- were not discussed in the paper [however VMPFC is difficult to image using fMRI because of susceptibility artifacts]. The somatic marker hypothesis is succinctly described by the title of one of Damasio's books: The Feeling of What Happens: Body and Emotion in the Making of Consciousness (1999).

Other components in the somatic marker circuit include the insular cortex, a region implicated in interoceptive awareness of bodily states (Craig, 2009), and somatosensory cortices responsive to external stimuli. Because activity in the anterior insula features primarily in the Twitter-warped distortion of the story, I'll start here with the authors' third hypothesis:
3. that activation in the anterior insula would peak and dissipate more quickly for CPP than for CSP or varieties of admiration.
In a way, this is a trivial prediction, because one can evaluate the sprained ankle narrative more quickly than the job loss scenario. In fact, I will argue below that simple behavioral response time might be a more precise measure of how long it takes to generate the emotion in question than is the hemodynamic response (blood flow changes, measured by the BOLD signal in fMRI) in the insular cortex. One reason for this is because of the significant delay (5-6 sec at least) between initial neural firing and the peak of the hemodynamic response, which is estimated using a procedure that is not trivial for something as complex as an emotional response (for a more detailed discussion of this issue, I recommend this PPT file from Jodi Culham's excellent fMRI 4 Newbies site).

Let's start with a simpler example. The figure below shows the averaged hemodynamic response function (HRF) in the primary visual cortex to a series of flashes. The HRF peaks at ~5 sec after the flash, whereas neurons in primary visual cortex fire within 50 msec and drop off shortly thereafter. Thus, the hemodynamic response to even a simple sensory stimulus lags behind neuronal firing by 5 sec.

Fig. 2 (Calhoun et al., 1998). Time courses from four regions in the calcarine cortex (Pl-P4) and the averaged response (CIRQ). Amplitude units are normalized to a maximum of one and a baseline of zero.

The next example shows the HRFs in occipital regions and the insula while subjects viewed rotating objects. The precise details aren't important here, but note the peak latency for the HRF in the insula is around 6-8 sec, with the later peak for novel objects (compared to repeated objects).

Fig 2C (Weigelt et al., 2007). Event-related deconvolved BOLD fMRI responses (GLM parameter estimates averaged across trials and subjects for all voxels in each ROI) reported against time for each of the experimental conditions.

That brings us back to Immordino-Yang et al. and the emotional narratives. In the figure below, note that the HRF time course does peak earlier for the CPP condition compared to the others, as predicted. However, the CSP condition rises at the same time, albeit with a later (very broad) peak.

Fig. 3 (Immordino-Yang et al., 2009). Event-related averages for the time courses of admiration and compassion in the anterior insula, with standard errors. Units are percentage change in BOLD signal and time in seconds; time courses are not corrected for hemodynamic delay. For display purposes, BOLD data have been linearly interpolated to 1-s resolution. The volume of interest is displayed in pink. Conditions: AV (green): admiration for virtue; AS (yellow): admiration for skill; CSP (blue): compassion for social pain; CPP (red): compassion for physical pain. Note the rapid rise and dissipation of CPP versus the slower and more sustained rise of CSP, AV, and AS.

It's critical to note that the onset of a felt emotion is not as easy to determine as the onset of a visual object. Although more detailed methods are in the Supplementary Materials not available as of this writing, it seems that respiration and heart rate data were obtained in 7 of the 13 subjects to help with this. I would say these psychophysiological responses, in concert with the participants' own reaction times for rating their subjective responses, would provide a more accurate measure of how long it takes to feel an emotion than the fMRI data. It's hard to know what an insular HRF of 6 sec vs. 10 sec means when watching a fast-paced movie or reading the CNN news crawl or yes, spending too much time on Twitter. Nonetheless, on the basis of these imprecise latency measures, the authors speculate:
If replicated, this finding could have important implications for the role of culture and education in the development and operation of social and moral systems; in order for emotions about the psychological situations of others to be induced and experienced, additional time may be needed for the introspective processing of culturally shaped social knowledge. The rapidity and parallel processing of attention-requiring information, which hallmark the digital age, might reduce the frequency of full experience of such emotions, with potentially negative consequences.
And there's your "Twitter is evil" angle.

I'll leave you with this final thought: where's the line between admiration and envy, between compassion and schadenfreude? There actually is a recent paper on the Neural Correlates of Envy and Schadenfreude (Takahashi et al., 2009), and for now I'll refer you to this nice summary in Pure Pedantry.

ADDENDUM (Monday 4:20 PM): You can read more details about the Methods in the Supporting Information, which is now freely available to all on the PNAS website, as is the open access article itself.

Recommended Twitter Reading:

Social media threats hyped by science reporting, not science (Ars Technica)

Experts say new scientific evidence helpfully justifies massive pre-existing moral prejudice. (Bad Science)

For the last time: that "Twitter is Evil" paper is not about Twitter! (Bioephemera)

Is Twitter evil? (Cosmic Log at MSNBC - despite the ridiculous headline, it's one of the few popular science articles to talk about the actual study... it even included a figure from the paper)

The Neurology of Twitter (The Neurocritic proposes an actual fMRI study of Twitter, complete with predicted results)

The Neurology of Twitter, Part 2 (Yet another recap of the media circus, with a time line of certain events)


Craig AD. How do you feel--now? The anterior insula and human awareness. (2009). Nat Rev Neurosci. 10:59-70.

Damasio AR. (1996). The somatic marker hypothesis and the possible functions of the prefrontal cortex. Philos Trans R Soc Lond B Biol Sci. 351:1413-20.

Mary Helen Immordino-Yang, Andrea McColl, Hanna Damasio, and Antonio Damasio. (2009). Neural correlates of admiration and compassion. Proceedings of the National Academy of Sciences.

Subscribe to Post Comments [Atom]


At April 20, 2009 2:13 PM, Anonymous Anonymous said...

Hi Neurocritic,

Nice article. Thanks for the link by the way. Did they operationalise any of the tests administered or were the narratives created ad hoc for the study? Did they use a self-rating scale for compassion? The reference to some of these emotions got me thinking - in your opinion what effect would the release of hormones associated with various emotions have on the imaging findings?



At April 20, 2009 4:01 PM, Blogger The Neurocritic said...

Hi Justin,

Thanks. In the Methods, the authors state that:

the narratives consisted of the experimenter’s previously scripted verbal account of each true story supplemented by materials adapted from various sources including television, the internet, documentary films, and radio. Narratives followed a consistent format (i.e., beginning, "This is the story of a woman/man who…" followed by a scripted story and the presentation of additional materials including video and/or still images of the protagonist. Each narrative required between 60 and 90 s to recount and finished with the experimenter asking, "How does this story make you feel?").The rating scale ranged from 1 to 4 where 1 was for "not particularly emotional," 2 was for "moderately," 3 was for "very strongly," and 4 was for "overwhelmingly emotional". More details are in the Supporting Information, which is now freely available to all on the PNAS website, as is the open access article.

Hormone levels weren't assessed in this study, but psychophysiological measures were. I'd expect that a myriad of hormones might influence the imaging findings, although the time frame for each trial was 18 sec... I don't know the duration of such responses. As for physiology, respiration rates did not differ between the conditions. Heart rates for AV, CSP, and CPP were higher than for AS and C.

Interestingly, Critchley and colleagues have examined patients with pure autonomic failure (PAF) to see how the lack of autonomic responses affects social, emotional, and motivational functioning. The PAF individuals showed an attenuation of empathy scores in one study (Chauhan et al., 2008), but they were unimpaired on tests of emotion recognition and theory of mind (Heims et al., 2004).

At April 21, 2009 6:55 AM, Anonymous bsci said...

You review of the time lags of hemodynamic responses misses a key factor. All those studies compare lag differences across regions. The Immordino-Yang study is looking at the task responses from a single region of interest. It is reasonable to assume that vascular dynamics within a voxel aren't task dependent thus any temporal differences are neural.

That said, the authors don't mention reaction time anywhere and that could affect the results. Also, it is possible that the temporal delay is happening in some other brain region that precedes the activation of the interior insula. Note that even reaction time differences ARE neural differences, but the differences might originate in earlier brain regions.

Reaction time measure and a more thorough whole-brain analysis of lags in significant brain regions could address both of these concerns.

At April 21, 2009 9:57 AM, Blogger The Neurocritic said...

My point in bringing up different brain regions in the time lag discussion is to initially raise the simpler issue that we're not talking about rapid-fire millisecond precision here, and that there's a built-in delay even for discrete visual stimuli in primary sensory cortex. The onset of an emotional response, I said, is not as clearly determined, so that compromises the within-region (or within voxel) comparison in the insula.

One must also take into account that the subjects were using the 5 sec cues to remember how they felt 1-2 hrs ago. I'm in complete agreement with you that RT measures should have been used and reported. And your suggestion of a whole-brain analysis of lags is a good one.

At April 21, 2009 12:23 PM, Anonymous Anonymous said...

Hi Neurocritic,

Thanks for your comments. I've just been following the thread and thought i'd just add the following
1. The authors seem to have done a lot of work in piloting the narratives they've used here. It would be useful for them to share the data from these pilots for further development and reuse as it seems quite a tall order to undertake the study at the same time as validating one of the measures.
2. I'm surprised that they're using retrospective evocation of emotions. Perhaps alternating narrative presentation with neutral prose might have allowed for within-subject control and 'real-time' assessment of emotional response. Each presentation could have been followed by the subject rating their feeling.
3. I might have misunderstood this but with whole brain analysis of lags does that mean that you can create a temporal sequence of activated regions within the brain. The results always seem to be presented as regions which have been significantly activated rather than pathways but this idea would make sense. I guess the difficulties would be in choosing thresholds for voxel firing and defining 'regions' within the pathway. However this would seem to be a neat way of analysing individual data and then combining it with other subjects. A pathway analysis might be complicated by multitasking where several pathways are being activated simultaneously.
4. Just a technical point but as the insular cortex is buried quite deep and has continuity with ACC how did the authors define the boundaries of the insular cortex? I've looked at a few studies where the approach for doing this seems to be lost in 'we used this software package'.



At May 01, 2009 9:53 AM, Anonymous Anonymous said...

I'm glad the authors used the phrase "if replicated, this finding...", because the chances of that happening are zilch (think voodoo correlations).


Post a Comment

<< Home

eXTReMe Tracker