Mechanisms

In basic terms, causal mechanisms can be characterized as space-time events that produce results. Mechanisms can be stable or unstable, singular or general. The mechanisms characteristic of sciences are most often stable and general, but unstable mechanisms are also scientific, as are singular ones.

The definition of causal mechanisms as events that produce results highlights their active dimension. In Glennan (1996, 52)'s words, ‘One cannot even identify a mechanism without saying what it is that the mechanism does'. Mechanisms are made up of entities ‘doing something', in other words, ‘a kettle boils, a bomb explodes, a hen lays an egg' (Glennan 2009, 327).^[140] In this way, entities produce or cause effects. A solar eclipse is an event produced by the alignment of the Moon and the Sun, due to the paths of their orbits. And this type of alignment causes eclipses of this kind.

The new mechanists agree on the role of entities, but they disagree on the nature of this ‘doing something'. There are two basic theories here, that of Glennan (1996, 2005, 2010a, 2011) and that of MDC (2000). Glennan holds that the activity of the entities is produced when they interact under certain conditions. This is a monist ontology: causal mechanisms produce results through the interaction of entities. The position put forth by MDC is dualist: the entities cause effects through the activity of these entities; therefore, mechanisms include entities and activities. The notion of activities refers to the behaviour of the parts of the mechanisms.^[141] Basi­cally the activities are the producers of effects, and therefore they are materialised causes.

The concept of interaction is hold in this work. Therefore, it is defended a mechanistic monist position in which diverse kinds of effects—biological, physical, chemical, neurological, social, etc.—are produced by the interaction of entities under specific conditions.^[142] Entities interacting under specific conditions are called causal mechanism. The steps involved in a complete description of a mechanism must be continuous; the description must be free of gaps that render specific steps unintelligible (Machamer 2004).

There are at least two ways to understand at causal mechanisms: as sequences of interconnected events that produce an effect, as Salmon (1984) argues, (the force of the wind produces a specific wave which sinks a boat), and as complex systems (which form part of a higher-level mechanism), the view held by Glennan (1996).^[143] In the latter case, the parts of a mechanism are lower-level mechanisms. Thus, for Glennan, an interaction involves a change in a property of one of the parts of the mechanism producing a change in a property of another part of this mechanism, which is understood as a complex system. Salmon sees causality as causal nexuses, while Glennan sees it as causal systems.

According to Salmon, a sequence of interconnected events might occur once or once in a while (fragile sequences), but a sequence may also repeat itself system­atically and invariably, as is the case, for example, with lunar eclipses (robust sequences).¹² This is due simply to the stability or robustness of the entities that interact and the regularity of these interactions. From my point of view, the robustness or stability of mechanisms do not require Glennan's systems approach; his approach also presents a problem of circularity (which he addresses by con­sidering mechanisms hierarchically).

In this paper an eclectic position is maintained, admitting that there are cases which mechanisms form part of a complex system, but this is not always so, and many mechanisms are interconnected sequences. It seems a good idea, under­standing, as Glennan does, that the parts of mechanisms are often mechanisms themselves. But we must also bear in mind that mechanisms are simple at some point. The dissolving of sugar in a liquid entails only the properties of the sugar crystals and those of the liquid. This is a case of a simple sequence of events made up of entities that interact to produce an effect. It does not represent an obstacle to establishing many different types of laws.

In both approaches, explaining phenomena involves providing a detailed exposition of how they were produced by interaction between entities. Causal analysis based on mechanisms involves knowing which entities are at work and the interactions between them. It is this knowledge which allows us to explain the occurrence of an event, as is the case of the death of the camellias Cartwright planted in her garden. On this occasion we have an explanation of a particular event, if we are able to establish the mechanism involved: the soil was poor in certain components, the temperature was too high, etc. The specific history leading from these entities' interactions to the death of the camellias explains the occur­rence of the event.

On the other hand, identifying mechanisms of a certain type allows us to explain general events, or to account for certain types of events such as cognitive disso­nance reduction and lunar eclipses. In this case, the causal chains that make up mechanisms are regularly activated given certain conditions, producing a certain type of effect. Thus, insofar as these mechanisms are neither unique nor unre­peatable, they can be identified as probabilistic, tendencies or even universal. A number of mechanisms as tendencies can be found in the social sciences: the

(Footnote 11 continued)

main positions on mechanisms, although there are other positions as well, for example that of Torres (2009).

2Glennan (2002) maintains this division in his systems approach to mechanisms.

‘forbidden fruit effect', the ‘spillover effect', the ‘compensation effect', ‘loss aversion', the ‘consumption effect' and ‘wishful thinking', among others.

An entity, a bomb that has blown up, for example, produces certain effects on another entity (a person's body) with which it interacts, producing specific effects (the person's death). In order to explain the effect produced, we must identify the mechanism that caused it (entities interacting). But, has the event been explained? Not quite, because in order to formulate a causal explanation we must determine how it was that those entities produced the effect they produced when they inter­acted. In other words, we must specify what makes the bomb capable of producing the effect it produced and what makes the person's body capable of reacting how it did to the explosion. To put it in general terms, what gives certain types of entities the capacity to produce the effects they produce when they interact? Not all interactions are causal. The answer to this question lies in the notion of ‘properties'.

4