The linguistic and cognitive sciences have severely underestimated the degree of linguistic diversity in the world. Part of the reason for this is that we have projected assumptions based on English and familiar languages onto the rest. We focus on some distortions this has introduced, especially in the study of semantics.

Are WEIRD societies unrepresentative of humanity? According to Henrich et al., they are not useful for generalizing about humans because they are at the extreme end of the distribution for societal formations. In their vision, it is best to stick with the “tried and true” traditional societies for speculations about human nature. This commentary offers a more realistic starting point, and, oddly enough, concludes that WEIRD populations may be more compatible with humans' evolved nature than are most traditional societies.

A literature in which most data are outliers is flawed, and the target article sounds a timely alarm call for the behavioural sciences. It also suggests remedies. We mostly concur, except for arguing that the importance of the fact that the researchers themselves are mostly outliers has been underplayed. Improving matters requires non-Western researchers, as well as research subjects.

We agree with Henrich et al. that documenting cultural universality and variability provides an indispensable window into human nature. We want to stress the mediating role development plays between evolution and culture. Moving beyond the mere documentation of universality or variability, developmental approaches can provide mechanistic explanations, linking ecology to phenotype. Combining phylogeny and adaptationism, evolutionary approaches can explain the properties of developmental systems.

Most psychological research consists of experiments that put people in artificial situations that elicit unnatural behavior whose ecological validity is unknown. Without knowing the psychocultural meaning of experimental situations, we cannot interpret the responses of WEIRD people, let alone people in other cultures. Psychology, like other sciences, needs to be solidly rooted in naturalistic observation and description of people around the world. Theory should be inductively developed and tested against real-world behavior.

Catchy acronyms such as “WEIRD population” are good mnemonics. However, they carry the danger of distracting us from deeper issues: how to sample populations, the validity of measuring instruments, the levels of processing involved. These need to be considered when assessing claims of universality regarding how the mind works “in general” – a dominant and highly rewarded drive in the behavioral and brain sciences.

Although North American undergraduates represent about 0.2% of humanity, and a very unrepresentative subset, they actually provide an advance look at what humanity is becoming. In the face of globalization, this is all the more reason to study the wonderful variants of the human condition before they become homogenized.

The objection, rightfully noted but then dismissed by Henrich et al., that the observed variation across populations “may be due to various methodological artifacts that arise from translating experiments across contexts” is a theoretically profound and potentially constructive criticism. It parallels Donald Campbell's concern that many cultural differences reported by psychologists “come from failures of communication misreported as differences.” Ironically, Campbell's doubt is a good foundation for investigations in cultural psychology.

From Plato to the present, philosophers have relied on intuitive judgments as evidence for or against philosophical theories. Most philosophers are WEIRD, highly educated, and male. The literature reviewed in the target article suggests that such people might have intuitions that differ from those of people in other groups. There is a growing body of evidence suggesting that they do.

In our response to the 28 (largely positive) commentaries from an esteemed collection of researchers, we (1) consolidate additional evidence, extensions, and amplifications offered by our commentators; (2) emphasize the value of integrating experimental and ethnographic methods, and show how researchers using behavioral games have done precisely this; (3) present our concerns with arguments from several commentators that separate variable “content” from “computations” or “basic processes”; (4) address concerns that the patterns we highlight marking WEIRD people as psychological outliers arise from aspects of the researchers and the research process; (5) respond to the claim that as members of the same species, humans must have the same invariant psychological processes; (6) address criticisms of our telescoping contrasts; and (7) return to the question of explaining why WEIRD people are psychologically unusual. We believe a broad-based behavioral science of human nature needs to integrate a variety of methods and apply them to diverse populations, well beyond the WEIRD samples it has largely relied upon.

The pivotal problem of comorbidity research lies in the psychometric foundation it rests on, that is, latent variable theory, in which a mental disorder is viewed as a latent variable that causes a constellation of symptoms. From this perspective, comorbidity is a (bi)directional relationship between multiple latent variables. We argue that such a latent variable perspective encounters serious problems in the study of comorbidity, and offer a radically different conceptualization in terms of a network approach, where comorbidity is hypothesized to arise from direct relations between symptoms of multiple disorders. We propose a method to visualize comorbidity networks and, based on an empirical network for major depression and generalized anxiety, we argue that this approach generates realistic hypotheses about pathways to comorbidity, overlapping symptoms, and diagnostic boundaries, that are not naturally accommodated by latent variable models: Some pathways to comorbidity through the symptom space are more likely than others; those pathways generally have the same direction (i.e., from symptoms of one disorder to symptoms of the other); overlapping symptoms play an important role in comorbidity; and boundaries between diagnostic categories are necessarily fuzzy.

We discuss the latent variables construct, particularly in regard to the following: that latent variables are considered as the sole explanatory factor of a disorder; that pragmatic concerns are ignored; and that the relationship of these variables to biological markers is not addressed. Further, we comment on the relationship between bridge symptoms and causality, and discuss the proposal in relationship to other constructs (endophenotypes, connectionist-inspired networks).

Although the network model represents a promising new approach to conceptualizing comorbidity in psychiatric diagnosis, the model applies most directly to Axis I symptom disorders; the degree to which the model generalizes to Axis II disorders remains open to question. This commentary addresses that issue, discussing opportunities and challenges in applying the network model to DSM-diagnosed personality pathology.

Network analysis is a promising step forward in efforts to align psychological assessment with explanatory theory in psychological science. The implications of Cramer et al.'s analysis are quite general. Networks analysis may illuminate functional relations not only among observable behaviors that comprise psychological disorders, but among cognitive and affective processes that causally contribute to everyday experience and action.

We agree with Cramer et al.'s goal of the discovery of causal relationships, but we argue that the authors' characterization of latent variable models (as deployed for such purposes) overlooks a wealth of extant possibilities. We provide a preliminary analysis of their data, using existing algorithms for causal inference and for the specification of latent variable models.

Psychological traits and disorders are often interrelated through shared genetic influences. A combination of maximum-likelihood structural equation modelling and multidimensional scaling enables us to open a window onto the genetic architecture at the symptom level, rather than at the level of latent genetic factors. We illustrate this approach using a study of cognitive abilities involving over 5,000 pairs of twins.

The network approach proposed by Cramer et al. suggests fascinating new directions of research on mental disorders. Research is needed to find evidence for the causal power of symptoms, to examine symptoms thoroughly, to investigate individual differences in edge strength, to discover etiological processes for each symptom, and to determine whether and why symptoms cohere into distinct mental disorders.

Cramer et al. make a good case for reconceptualizing comorbid psychopathologies in terms of complex network theory. We suggest the need for an extension of their network model to include reference to latent causes. We also draw attention to a neglected approach to theory appraisal that might usefully be incorporated into the methodology of network theory.

The network approach to psychiatric phenomena has the potential to clarify and enhance psychiatric diagnosis and classification. However, its generally well-justified anti-essentialism views psychiatric disorders as invariably fuzzy and arbitrary, and overlooks the likelihood that the domain includes some latent categories. Network models misrepresent these categories, and fail to recognize that some comorbidity may represent valid co-occurrence of discrete conditions.

Cramer et al.'s account of comorbidity comes with a substantive philosophical view concerning the nature of psychological disorders. Although the network account is responsive to problems with extant approaches, it faces several practical and conceptual challenges of its own, especially in cases where the individual differences in network structures require the analysis of intra-individual time-series data.