One of the key questions at the heart of Heckers’s reflections on the uses of neuroimaging in psychiatry is whether the actual uses to which neuroimaging are being put in experimental research on the neural markers and mechanisms of psychiatric disorders have application in clinical psychiatry. He points to a mismatch in goals and methods that, it seems to me, demands further scrutiny, and scrutiny of a sort that philosophers might usefully supply. In addition to the questions of causal inference, to which I direct my attention immediately in what follows, are important questions about the difference between general truths about the brains of people with schizophrenia and truths about specific, single cases. Even if the tools of psychiatric research succeed in gleaning causal information about psychiatric disorders in general terms, there will always be an inductive leap in relating the general to the particular, in seeing the individual as properly both an example of the generalizations discovered in research on populations (and controls) and, consequently, as a fitting recipient of the predictions and explanations those generalizations deliver. This gap between the general and the particular lurks in the background of Heckers’s important discussion, but perhaps could be given more of a central place, and would help us distinguish which problems attend the causal uses of neuroimaging and which attend the uniquely messy business of drawing conclusions about single-case causes on the basis of causal knowledge that holds, only for the most part, in certain conditions that we cannot fully adumbrate.

There has been a lot of interest lately in the possibility of downward causation, the idea that higher-level processes or events can cause lower-level ones. Downward causation, if it could be made sense of, is seen by some as a way to approach the philosophical problem of free will naturalistically, while others have heralded it as a promising way to explain aspects of the pathogenesis or treatment of psychiatric conditions. But the possibility of downward causation has been called into question because it is claimed to require causation across levels, and interlevel causation has been viewed as problematic. Reductionist accounts, for example, hold that entities or processes at higher levels can be reduced to entities or processes at lower levels, and that causal relations only “really” occur at the most fundamental level. On this basis, Kim’s (1998) causal exclusion argument holds that mental states do not enter causal relations and are merely epiphenomenal. Alternatively, the New Mechanist account holds that neuroscientific explanation is mechanistic explanation, and that the relationship between mechanisms at different levels is one of constitution (part/whole). Because constitution relations are not causal relations, interlevel causation does not exist: causation can only occur within a level (see, e.g., Bechtel & Craver, 2007; Craver, 2007). Several thinkers have suggested that if we can rid ourselves of levels, these problems and others would disappear (Eronen, 2013).

Carl Craver begins his chapter by noting that in psychiatry, mechanistic models attempt to understand the causal factors that produce, underlie, or maintain disorders. Knowledge of mechanisms, therefore, would support the goals of preventing, ameliorating, and even curing psychiatric disorders and thus represent the discipline’s highest aspirations for psychiatric explanation. Some thinkers, however, are concerned that under the banner of mechanism resides a static view of phenomenon that does not bear in mind that things constantly change. Another perceived problem with a mechanistic approach is a reductionism that dismisses higher-level causal structures in favor of using new technologies designed primarily to study the smallest things. This chapter argues that such concerns are not applicable to the new mechanists in the philosophy of science.

The interventionist approach to causation has become a dominant perspective in the philosophy of science, for many reasons. One reason is that it can be generalized to causation across the physical sciences, biological sciences, psychological sciences, and social sciences without losing plausibility; in other words, in an interventionist framework, causality is not “softer” in psychology than it is in the physical sciences. That is not to say that the applications work the same way across sciences. For instance, in their chapter, Roskies, Busch, and Walton note that applying a neuroscientific intervention directly to the interconnected causal network that constitutes the brain is laden with complications.

Jim Woodward has, for a number of years, been a leading philosophical theorist of causation, playing a particularly prominent role in the articulation of the influential theory of interventionism as outlined in his book, Making Things Happen (Woodward, 2003). Readers are therefore in for a treat in this essay (Chapter 2) in which he reviews his current thinking about causation as it applies to psychiatric research and more particularly to the clarification of the impact of environmental risk factors by epidemiological methods. In this introduction, I will seek only to expand on a modest number of points among the many he covers in this review. Most of my points are focused on one relatively well researched risk factor for psychiatric disorders – childhood sexual abuse.

This chapter explores the idea that the concept of dynamical gestalt, defined across a heterogeneous collection of interacting variables/factors/processes (including biopsychosocial factors) and distributed across different scales, can provide a level-free explanation of psychiatric disorders. In this view, all relevant factors are defined by their causal interaction, albeit at different temporal and spatial scales. Accordingly, there is no reason to think that social factors are on a higher level than neuronal processes, or that neural processes are more basic than social factors in the etiology of psychiatric disorders. In combination with the model of interventionist causality, this approach allows for the development of a productive way of explaining psychiatric disorders and for insight into therapeutic practices.

The Descent of Man sparked many scientific revolutions, but none were more intellectually profound or ethically challenging than the exposure of human behavior to the bright light of natural science. Before Darwin, people could still hold out some hope that we were separated from the rest of nature by our souls, our creation in the image of God. Human behavior had always been in the domain of theology, or, among skeptics, philosophy; “behavioral science” didn’t exist. After Darwin, scientifically clear-eyed people had to recognize that behavior – personality, intellect, family, culture – evolved just as surely as our bipedal posture. No one was more attuned to this revolution than Darwin’s cousin, Francis Galton. Galton made the application of Darwin’s new science to human behavior, in all its diversity, his life’s work. Human anatomy had been naturalized once scientific dissection of cadavers was permitted; the psychologist Wilhelm Wundt was already naturalizing human perception; Galton possessed unwavering confidence that the rest of human behavior was not far behind.

In their chapter, Federica Russo and Dingmar van Eck offer a novel approach to the network theory of mental disorders. According to the network theory, a syndrome such as depression occurs when the symptoms of depression become organized into an integrated causal network that remains stable over time. Rather than explaining the occurrence of depression with respect to hidden, underlying causes, the network theory emphasizes direct casual relationships between the symptoms themselves. According to this view, symptoms are not manifestations of an underlying common cause, but parts of the disorder. The organizational structure of the symptoms, therefore, constitutes the disorder. Russo and van Eck also suggest that as an alternative to the underlying cause model, the explanatory resources of the network model are underdeveloped. The purpose of their chapter is to expand on the explanatory potential of network models.

John Campbell reviews for us how to think about the causal processes that might be at work in psychiatry. The laws of physics cannot be a plausible model because it deals with matter – particles, planets, and the like. In psychiatry, we deal with patients who are both brained and minded. The brain is a physical system but one of staggering complexity. The mind is a whole different kind of thing, and it is here that empathic understanding must come into play. Given now important our interactions with our conspecifics are – our spouses, children, friends, and particularly enemies – evolution has likely tuned our emphatic understanding for accuracy to help us intuit the motivations and internal emotional states of our fellow humans.

After a brief overview, Lauren N. Ross and Kenenth S. Kendler begin their chapter reviewing the history of the proximal versus distal cause distinction in public health and epidemiology. Its purpose is to provide a framework for understanding how social and biological causes jointly impact the development of disease and health outcomes. This distinction traditionally utilizes a causal pathway model in which, for example, upstream, social factors such as organizational policies for remote work influence downstream, direct biological factors such as infection.

For the last 150 years, schizophrenia has been considered a brain disease leading to a cognitive impairment. A widespread representation of schizophrenia considers it a form of mental deficit. This chapter evaluates to what extent schizophrenia corresponds to a standard folk-psychological notion of disease (onset, pathological process, insight into illness, etc.). Parnas and Zandersen present examples of phenomenological characteristics of schizophrenia that question its status as a typical disease. Specifically, they portray the affliction of the subject (self-disorder), an affliction that prevents a full insight into illness. Self-disorders typically begin during childhood and adolescence, and dating the onset of schizophrenia becomes a conceptual rather than an empirical issue. Next, the authors discuss a characteristic configuration of psychosis, namely the phenomenon of “double bookkeeping” that implies a persistent parallel double orientation to both a social reality and a private psychotic world. Finally, they look at some characteristics of course and outcome that do not conform to a standard notion of disease. The authors conclude that to progress in our scientific research, psychiatry needs to refocus its interest on the issues of subjectivity and psychological processes involved in schizophrenia.