Machresis and Mediated Downward Causation
Elsewhere I have defended the coherence of machresis, and the S-emergence it underpins, and also rebutted an array of objections to its possibility and actuality.4 Given my purposes here, let me instead simply focus on drawing out how we still have a species of downward causation when we have machretic relations. Just as Laughlin outlines in the passage quoted in the introduction, taking FDR to be machretic means that when we have S-emergence we have two species of ‘vertical’ non-causal determination relations. A S-emergent property is composed—and hence ‘upwardly’ non-causally determined—by certain lower-level entities. But these components have some of their powers only as a result of the ‘downward’ machretic determination of the S-emergent composed property instance. At a time, due to this pair of non-causal determination relations, we therefore have mutually determinative entities and, as Laughlin implies, two determinative vectors in nature.
Against this background, we have an array of causal relations whose nature I now want to draw out. To this end, it is helpful if we make a distinction between two species of “downward causation”. Let us the term “direct” downward causation to refer to a composed entity that directly causally acts upon its own components which is obviously the type of downward causation posited by those who take FDR to be a causal relation. And let us use the term “mediated” downward causation to refer to a composed entity that bears a causal relation to lower-level entities that are not its components and where these lower-level entities are effects of components of the composed entity.
With this distinction in hand, let us assume we have the S-emergent instance G that machretically determines that some of its realizer property instances P1-Pn contribute differential powers that have instances of lower-level property Pz as their effects. Let us assume that G is instantiated in an individual s* that has as components individuals s1-sn which instantiate P1-Pn. And let us also assume that, given the powers contributed by G to s*, s* produces instances of property H in other higher-level individuals. In addition, let us take it to be the case that, given their powers, P1-Pn cause instances of Pz at the lower level as the effect of their differential and other powers, where P1-Pn would not have the differential powers if G had been absent.
Scientific Emergentism and Its Move beyond 259
This situation is framed in Figure 14.1, where the enclosed, straight vertical arrows upwards are compositional relations at certain times; enclosed, curved vertical arrows downwards are machretic relations; and all the solid arrows are causal relations.
Notice that Figure 14.1 includes both horizontal and also diagonal causal relations, but no vertical causal arrows. This is no accident, so let us work through the reasons already outlined about why this is situation. The mutual determinativity of the S-emergent composed entity, and its components, means that these entities are joint causes of both higher- and lower-level causal effects—hence the downward ‘diagonal’ causal arrow from G to Pz-Pzn. For instance, the S-emergent property G in my scenario is plausibly a joint cause, together with the instances of realizers like P1, of effects involving instances of Pz. But we do not have any direct downward causation, since for the reasons outlined in earlier sections the emergent realized property G cannot cause its realizers, like P1, to contribute certain differential powers—G only machretically determines P1 contributes these powers. Instead, P1 and other realizers causally interact with certain lower-level entities through these differential powers and G is thus also a cause, albeit a mediated downward cause, of such effects. And we will plausibly have counterfactual and manipulability relations between G and these effects in instances of Pz, since there are ideal interventions on G with respect to these effects.
Under the scientific emergentist picture endorsing machresis as FDR, downward causation will therefore be as pervasive as S-emergence itself, but this is only mediated, diagonal downward causation. Furthermore, such mediated, diagonal downward causation fits nicely with the kinds of downward causal relations illuminated by application of interventionist frameworks—for such causal relations do not hold between composed entities and their own components which we have seen cannot bear manip- ulability relations. As well as machresis providing a coherent picture for S-emergence that accommodates its key features, we therefore also have good reasons to believe that machresis makes sense of the only kind of downward causation that plausibly exists in science and nature.