Gaze patterns and brain activations in humans and marmosets in the Frith-Happé theory-of-mind animation task
Curation statements for this article:-
Curated by eLife
eLife assessment
This study provides useful findings regarding the capacity for mental state attribution on the Frith-Happé task in a highly social non-human primate species, the marmoset. The methods are solid, integrating validated brain imaging and eye-tracking techniques, however, the theoretical analysis is incomplete due to the omission of a "goal-directed condition." The inclusion of the goal-directed condition would enable a stronger interpretation of the findings and would be of broad interest to neuroscientists working in social and affective sciences.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
Abstract
Theory of Mind (ToM) refers to the cognitive ability to attribute mental states to other individuals. This ability extends even to the attribution of mental states to animations featuring simple geometric shapes, such as the Frith-Happé animations in which two triangles move either purposelessly (Random condition), exhibit purely physical movement (Goal-directed condition), or move as if one triangle is reacting to the other triangle’s mental states (ToM condition). While this capacity in humans has been thoroughly established, research on nonhuman primates has yielded inconsistent results. This study explored how marmosets ( Callithrix jacchus ), a highly social primate species, process Frith-Happé animations by examining gaze patterns and brain activations of marmosets and humans as they observed these animations. We revealed that both marmosets and humans exhibited longer fixations on one of the triangles in ToM animations, compared to other conditions. However, we did not observe the same pattern of longer overall fixation duration on the ToM animations in marmosets as identified in humans. Furthermore, our findings reveal that both species activated extensive and comparable brain networks when viewing ToM versus Random animations, suggesting that marmosets differentiate between these scenarios similarly to humans. While marmosets did not mimic human overall fixation patterns, their gaze behavior and neural activations indicate a distinction between ToM and non-ToM scenarios. This study expands our understanding of nonhuman primate cognitive abilities, shedding light on potential similarities and differences in ToM processing between marmosets and humans.
Article activity feed
-
-
eLife assessment
This study provides useful findings regarding the capacity for mental state attribution on the Frith-Happé task in a highly social non-human primate species, the marmoset. The methods are solid, integrating validated brain imaging and eye-tracking techniques, however, the theoretical analysis is incomplete due to the omission of a "goal-directed condition." The inclusion of the goal-directed condition would enable a stronger interpretation of the findings and would be of broad interest to neuroscientists working in social and affective sciences.
-
Reviewer #1 (Public Review):
This is an interesting study deploying convergent methodologies to address a timely question: can non-human primates distinguish theory of mind from random behaviours during passive viewing of animated shapes, and what brain regions are implicated? As the authors note, fMRI studies of brain activation in response to the theory of mind stimuli in non-human primates are scarce, and none have explored the processing of abstract stimuli in this context.
The major strengths of the study are the application of the Frith-Happé shapes task in a group of marmosets during fMRI in conjunction with concurrent eye tracking recording. Eye tracking is a very nice addition as it enables the authors to determine the gaze patterns and fixation duration on distinct aspects of the task stimuli (e.g., large triangle versus small …
Reviewer #1 (Public Review):
This is an interesting study deploying convergent methodologies to address a timely question: can non-human primates distinguish theory of mind from random behaviours during passive viewing of animated shapes, and what brain regions are implicated? As the authors note, fMRI studies of brain activation in response to the theory of mind stimuli in non-human primates are scarce, and none have explored the processing of abstract stimuli in this context.
The major strengths of the study are the application of the Frith-Happé shapes task in a group of marmosets during fMRI in conjunction with concurrent eye tracking recording. Eye tracking is a very nice addition as it enables the authors to determine the gaze patterns and fixation duration on distinct aspects of the task stimuli (e.g., large triangle versus small triangle) as well as group differences. Overall, the study seems well-designed and technically rigorous, and the major conclusions appear to be supported by the data.
However, there is one aspect I would appreciate some clarity on, namely the failure to include the original "Goal directed condition" from the Frith-Happé task. The authors contrast visuo-oculomotor and fMRI activation between the Random (no discernible interaction or purposeful behaviour) and the ToM (goal-directed behaviours with mental interaction) but neglect the intermediate step of physical interaction between the shapes that the Goal-directed behaviour condition portrays. As such, it is difficult to make clear statements as to what the activation patterns in the ToM condition represent - perhaps this merely reflects the processing of an unfolding narrative rather than random movements.
-
Reviewer #2 (Public Review):
In this study, Dureux and colleagues show that marmosets are sensitive to the Frith and Happe social illusion. This result is particularly interesting from an evolutionary perspective as rhesus macaques are insensitive to this social illusion.
Although marmosets show sensitivity to the illusion of social interaction between two geometric shapes, behavioural and neuronal evidence also show differences between humans and marmosets.
-
Reviewer #3 (Public Review):
To assess the degree to which highly social primates like marmosets share a human-like Theory of Mind (ToM), the authors used eye tracking and functional magnetic resonance brain imaging on marmosets and humans who were viewing two of the three categories from classic Frith-Happé animations. Humans viewing the ToM animations showed, relative to the random movement animations, longer fixation times, more viewing of the large shape, and more viewing of the small shape. In contrast, the marmosets did not differ in their viewing of the ToM videos as a category and did not show differential viewing of the small shape. The marmosets did show differential viewing of the large shape, but this difference was blunted relative to that seen in humans. Neurally, both species showed widespread brain activation in many …
Reviewer #3 (Public Review):
To assess the degree to which highly social primates like marmosets share a human-like Theory of Mind (ToM), the authors used eye tracking and functional magnetic resonance brain imaging on marmosets and humans who were viewing two of the three categories from classic Frith-Happé animations. Humans viewing the ToM animations showed, relative to the random movement animations, longer fixation times, more viewing of the large shape, and more viewing of the small shape. In contrast, the marmosets did not differ in their viewing of the ToM videos as a category and did not show differential viewing of the small shape. The marmosets did show differential viewing of the large shape, but this difference was blunted relative to that seen in humans. Neurally, both species showed widespread brain activation in many areas that discriminated between ToM videos and random movement videos. This pattern of activation partially overlapped and partially was different in humans and marmosets. It was also partially overlapping and partially different when comparing humans in this study to humans in another study. Overall, the authors conclude that their evidence cannot address whether marmosets have a Theory of Mind, but that marmosets show a "clear preference for interacting shapes" that may be an ancestral form of human Theory of Mind.
There are several laudable strengths to this report. It reports a direct human/monkey comparison. It uses a robust population of subjects, especially for the monkey experiment. It uses strong imaging methods that use modern parcellation maps, compares human data from this study to comparable data from another study, and accounts for lateralization differences convincingly using maps of signal-to-noise ratio. It uses eye-tracking methods and stimuli that are solidly grounded in the human literature and that has recently been used in a different monkey species.
Unfortunately, the weaknesses of this report limit its interpretability. First, it omits one of the three major categories of the Frith-Happé animations: Goal-Directed actions. Data from this category are critical because they provide a case where the shapes are engaging in biological motion but are not behaving as if they attribute minds to each other. Without including it, readers cannot interpret whether any given finding is due to biological motion or mentalizing. Second, the study did not gather explicit reports of mental state attribution from humans. This does not allow for a manipulation check about whether humans were even engaging in mentalizing and does not allow the researchers to separate out what brain activation patterns are due to mentalizing and which are due to eye movements or stimulus movement. Third, in interpreting the data, the researchers gloss over the major species differences and primarily focus on one small species similarity. Both this study and a previous human study (Klein et al., 2009, Quart. J. Exp. Psychol.) have shown longer fixations for the ToM videos relative to the random motion videos and that these fixations correlate with explicit ratings of the intentionality of the shapes (Klein et al., 2009). That the marmosets don't show this difference should be a major piece of evidence against the hypothesis that they are engaging in anything like mentalizing. The marmosets also failed to show a viewing difference for the small shape. In short, the small viewing difference in the large shape, itself blunted relative to that seen in humans, is not sufficient evidence to justify the conclusion that marmosets engage in anything like ToM or even that they show a "clear preference for interacting shapes". Fourth, alternative explanations for the small differences that do exist were not sufficiently explored. The videos that make up the categories in the Frith-Happé animations differ in many ways, such as in the amount of visual motion, smoothness/jerkiness of motion, amount of the screen taken up by shapes vs white space, etc. Indeed, in the prior study to use these stimuli with monkeys, the authors also found that the categories differed in viewing parameters but that this difference disappeared once low-level visual motion was accounted for (Schafroth et al., 2021, Sci. Rep.). Without a similar analysis here or a second experiment that assesses generalization to stimuli that don't differ on low-level perceptual features, readers cannot know whether the small viewing difference that exists is due to something like mentalizing or something about low-level visual motion. Indeed, other studies have found overlapping brain activity patterns in monkeys that are driven primarily by low-level visual motion (e.g., Russ et al., 2015, Neuroimage). Fifth, the prior monkey study to use these stimuli raised the point that these stimuli may not even be appropriate to test ToM in nonhumans. Human-like displays of "mocking", "coaxing", or "seducing" are likely meaningless to monkeys. This weakness has not been addressed in the current study.
Considering the weaknesses in the behavioral methods, the well-collected neural activity patterns cannot be interpreted in a meaningful way. As such, the authors' conclusions are not justified at the current time. Nevertheless, this report may be useful to others who attempt similar experiments of their own.
-
