Osiurak & Reynaud (O&R) detail a two-pronged case for technical-reasoning skills as primary in cumulative technological culture (CTC), looking to neurophysiological research on the one hand, and micro-society paradigms on the other. Our commentary will focus on the latter line of research.
In the micro-society paradigm, participants are tasked with building an artifact (e.g., a wire tower or paper airplane) within an allotted period. These artifacts (one for each participant) are then assessed on dimensions such as attractiveness, size, and functionality. Critically, participants complete the task as members of a “chain,” such that upon completion of the task, the next participant in the chain completes the task as well. Participants either simply observe the actions of others in their chain, or are allowed to communicate with them before attempting to build the artifact on their own. Of interest are relationships that might emerge between various cognitive capacities and both individual and cumulative performance under these two conditions. O&R's key conclusion from this research is that technical-reasoning skills predict cumulative performance better than social-reasoning skills.
In this research, technical-reasoning skills were assessed solely via participants' performance on two subsets of the Batterie Multifactorielle d'Aptitudes (NV7 battery), measuring ability to select appropriate tools for a given task and mental rotation. Similarly, social-reasoning skills were assessed via just two tasks: the Reading-the-Mind-in-the-Eyes Task (RMET), measuring ability to label emotions based on facial expressions, and the Comic Strip (CS) task, measuring ability to infer an agent's future action based on prior contextual information.
There are several reasons for questioning O&R's findings as an adequate foundation for their inference that technical-reasoning skills account for CTC to a greater degree than social-reasoning skills. First, the only two measures O&R employed to assess the role of social reasoning in CTC are unlikely to index individual variability in the types of social-reasoning skills that are argued to facilitate CTC (see Tomasello Reference Tomasello2016). Specifically, the RMET indexes only the ability to accurately label a decontextualized facial emotional expression (see Aviezer et al. Reference Aviezer, Ensenberg and Hassin2017); how this ability relates to general social understanding is not clear (see Oakley et al. Reference Oakley, Brewer, Bird and Catmur2016). As well, a recent meta-analysis of the psychometric properties of the CS revealed a negatively skewed pattern suggesting possible ceiling effects; it may not be well-suited for investigating individual differences (Davidson et al. Reference Davidson, Lesser, Parente and Fiszdon2018), and thus is likely inappropriate for assessing relationships between social reasoning and CTC.
Second, flawed logic affects O&R's analysis: just because individual differences on particular social-reasoning tasks (e.g., RMET and CS) fail to predict outcomes (e.g., cumulative performance) doesn't mean those (or other) social-reasoning skills are not critical to the achievement of those outcomes. Key aspects of the skill at issue might be quite central to the outcome; aspects of the skill that happen to display individual differences may be what are irrelevant. To illustrate, consider a scenario in which researchers observe no correlation between performance on a particular working memory task and vocabulary size. It would be misguided to conclude from such a finding that working memory more generally is unrelated to language acquisition. Relatedly, a more general point: O&R's case is problematic because it involves drawing a strong inference – social-reasoning skills aren't key to CTC – from a null finding (absence of a particular correlation).
Third, the micro-society paradigm, although innovative in its use of tangible dependent variables, does not adequately capture the cultural, historical, and ecological circumstances underlying CTC in the real world. Ironically, O&R contrast their views with those of Boyd et al.'s (Reference Boyd, Richerson and Henrich2011) “cultural niche” approach, yet Boyd and colleagues explicitly refer to adaptation (e.g., sophisticated hunting or shelter building methods) whereas the artifacts in O&R's tasks (paper airplane and tower) do not serve what would normally be regarded as adaptive functions. Without genuine adaptive pressure, there may be little communal interest in achieving cumulative progress. Therefore, O&R's micro-society findings may radically underestimate the role of social-reasoning factors in CTC.
Fourth, the micro-society paradigm is biased toward an individual-centered cognitivist approach because the constructed artifacts are always built by individuals who are alone in a room. Collaboration, an important component of the culture-centered approach to CTC (see Tomasello Reference Tomasello2016), is thus not observable because of a task design that prohibits social interaction during the actual construction phase.
Finally, O&R's conceptual analysis lacks precision. Specifically, they employ terms such as technical reasoning, mechanical knowledge, and causal understanding interchangeably to refer to the proposed primary “driver” of CTC, even though these terms are arguably distinguishable in meaning. Furthermore, the scope of O&R's claim is not clear. In other places (e.g., De Oliveira et al. Reference De Oliveira, Reynaud and Osiurak2019), the authors limit their claims to technology to the exclusion of other forms of cumulative culture such as music (itself a distinction that is not self-evident), but such clarification is missing in the target article, generating genuine theoretical confusion. Finally, although O&R describe theory-of-mind understanding as orthogonal to causal reasoning, it is important to note that causal reasoning is argued by many to be a defining feature of theory-of-mind development (e.g., see Penn et al. Reference Penn, Holyoak and Povinelli2008b). Pitting the two against one another is potentially misleading and theoretically problematic.
Similar to others (e.g., Gopnik & Wellman Reference Gopnik, Wellman, Hirschfeld and Gelman1994; Penn & Povinelli Reference Penn and Povinelli2007; Wellman & Gelman Reference Wellman and Gelman1992), we find it highly plausible that causal learning is a primary engine behind humans' acquisition of both sophisticated social- and technical-reasoning skills. Furthermore, despite the lack of evidence, we agree with O&R that causal understanding likely plays an important role in CTC with language – another inter-related skill-set – operating as a powerful catalyst (Sterelny Reference Sterelny2016). Moving forward, we suggest that future research using the micro-society paradigm introduce both adaptational and collaborative components to achieve greater ecological validity and provide a more comprehensive view of the dynamics of CTC.
Osiurak & Reynaud (O&R) detail a two-pronged case for technical-reasoning skills as primary in cumulative technological culture (CTC), looking to neurophysiological research on the one hand, and micro-society paradigms on the other. Our commentary will focus on the latter line of research.
In the micro-society paradigm, participants are tasked with building an artifact (e.g., a wire tower or paper airplane) within an allotted period. These artifacts (one for each participant) are then assessed on dimensions such as attractiveness, size, and functionality. Critically, participants complete the task as members of a “chain,” such that upon completion of the task, the next participant in the chain completes the task as well. Participants either simply observe the actions of others in their chain, or are allowed to communicate with them before attempting to build the artifact on their own. Of interest are relationships that might emerge between various cognitive capacities and both individual and cumulative performance under these two conditions. O&R's key conclusion from this research is that technical-reasoning skills predict cumulative performance better than social-reasoning skills.
In this research, technical-reasoning skills were assessed solely via participants' performance on two subsets of the Batterie Multifactorielle d'Aptitudes (NV7 battery), measuring ability to select appropriate tools for a given task and mental rotation. Similarly, social-reasoning skills were assessed via just two tasks: the Reading-the-Mind-in-the-Eyes Task (RMET), measuring ability to label emotions based on facial expressions, and the Comic Strip (CS) task, measuring ability to infer an agent's future action based on prior contextual information.
There are several reasons for questioning O&R's findings as an adequate foundation for their inference that technical-reasoning skills account for CTC to a greater degree than social-reasoning skills. First, the only two measures O&R employed to assess the role of social reasoning in CTC are unlikely to index individual variability in the types of social-reasoning skills that are argued to facilitate CTC (see Tomasello Reference Tomasello2016). Specifically, the RMET indexes only the ability to accurately label a decontextualized facial emotional expression (see Aviezer et al. Reference Aviezer, Ensenberg and Hassin2017); how this ability relates to general social understanding is not clear (see Oakley et al. Reference Oakley, Brewer, Bird and Catmur2016). As well, a recent meta-analysis of the psychometric properties of the CS revealed a negatively skewed pattern suggesting possible ceiling effects; it may not be well-suited for investigating individual differences (Davidson et al. Reference Davidson, Lesser, Parente and Fiszdon2018), and thus is likely inappropriate for assessing relationships between social reasoning and CTC.
Second, flawed logic affects O&R's analysis: just because individual differences on particular social-reasoning tasks (e.g., RMET and CS) fail to predict outcomes (e.g., cumulative performance) doesn't mean those (or other) social-reasoning skills are not critical to the achievement of those outcomes. Key aspects of the skill at issue might be quite central to the outcome; aspects of the skill that happen to display individual differences may be what are irrelevant. To illustrate, consider a scenario in which researchers observe no correlation between performance on a particular working memory task and vocabulary size. It would be misguided to conclude from such a finding that working memory more generally is unrelated to language acquisition. Relatedly, a more general point: O&R's case is problematic because it involves drawing a strong inference – social-reasoning skills aren't key to CTC – from a null finding (absence of a particular correlation).
Third, the micro-society paradigm, although innovative in its use of tangible dependent variables, does not adequately capture the cultural, historical, and ecological circumstances underlying CTC in the real world. Ironically, O&R contrast their views with those of Boyd et al.'s (Reference Boyd, Richerson and Henrich2011) “cultural niche” approach, yet Boyd and colleagues explicitly refer to adaptation (e.g., sophisticated hunting or shelter building methods) whereas the artifacts in O&R's tasks (paper airplane and tower) do not serve what would normally be regarded as adaptive functions. Without genuine adaptive pressure, there may be little communal interest in achieving cumulative progress. Therefore, O&R's micro-society findings may radically underestimate the role of social-reasoning factors in CTC.
Fourth, the micro-society paradigm is biased toward an individual-centered cognitivist approach because the constructed artifacts are always built by individuals who are alone in a room. Collaboration, an important component of the culture-centered approach to CTC (see Tomasello Reference Tomasello2016), is thus not observable because of a task design that prohibits social interaction during the actual construction phase.
Finally, O&R's conceptual analysis lacks precision. Specifically, they employ terms such as technical reasoning, mechanical knowledge, and causal understanding interchangeably to refer to the proposed primary “driver” of CTC, even though these terms are arguably distinguishable in meaning. Furthermore, the scope of O&R's claim is not clear. In other places (e.g., De Oliveira et al. Reference De Oliveira, Reynaud and Osiurak2019), the authors limit their claims to technology to the exclusion of other forms of cumulative culture such as music (itself a distinction that is not self-evident), but such clarification is missing in the target article, generating genuine theoretical confusion. Finally, although O&R describe theory-of-mind understanding as orthogonal to causal reasoning, it is important to note that causal reasoning is argued by many to be a defining feature of theory-of-mind development (e.g., see Penn et al. Reference Penn, Holyoak and Povinelli2008b). Pitting the two against one another is potentially misleading and theoretically problematic.
Similar to others (e.g., Gopnik & Wellman Reference Gopnik, Wellman, Hirschfeld and Gelman1994; Penn & Povinelli Reference Penn and Povinelli2007; Wellman & Gelman Reference Wellman and Gelman1992), we find it highly plausible that causal learning is a primary engine behind humans' acquisition of both sophisticated social- and technical-reasoning skills. Furthermore, despite the lack of evidence, we agree with O&R that causal understanding likely plays an important role in CTC with language – another inter-related skill-set – operating as a powerful catalyst (Sterelny Reference Sterelny2016). Moving forward, we suggest that future research using the micro-society paradigm introduce both adaptational and collaborative components to achieve greater ecological validity and provide a more comprehensive view of the dynamics of CTC.
Financial support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Conflict of interest
None.