Neuroscience, as a discipline rooted in the biological sciences, produces explanations based on mechanisms. In general, a mechanism implies the recognition of relevant physical entities or components and of their modes of causal interaction. When at work, it is this (usually non-observable) interaction which produces a given (observable) phenomenon. Most, if not all, of biological processes, are describable through mechanisms: photosynthesis, respiration, hormone action, gene expression, homeostasis maintenance, organ functioning, neural firing, sensory transduction, and the list goes on. For example, we can observe through the recording of electrical activity the firing of action potentials by neurons. This phenomenon, a change in voltage that rapidly departs from and returns to baseline, it is explained in terms of interactions between chemical components. The cell membrane contains molecular pores, called channels, which open and close depending on the membrane's voltage. If open, channels let ions pass and cross the membrane, changing, in turn, the membrane's voltage. The voltage sensitivity and time course of the different channels, the ion selectivity of the channels, the ionic concentrations on each side of the membrane, and the ion's charge, through their interplay, explain the firing of action potentials. The example also illustrates that mechanisms operate at given levels, which change according to the scientist's explanatory goals. The firing of action potentials constitutes a cellular-level phenomenon, and its mechanism runs through both cellular and sub-cellular levels. Other mechanisms -such as those relevant to cognitive phenomena- can occur at higher levels. In this case, the relevant entities could be brain nuclei or areas, i.e. a tissue-level entity. The causal interaction would be given by changes in electrical activity originated in one area and producing neurotransmitter release in a different one. Multiple brain regions, with reciprocal influence through this kind of activity, are probably important mechanisms for cognitive phenomena.
Philosophers of biology and neuroscience have articulated detailed and sophisticated interpretations of the modes of explanation of neuroscientists associated with mechanism-based research \cite{Machamer_2000,Bechtel_2011,Bechtel2008,Craver2007}.
Most neuroscientific mechanisms, regardless of the level at which they are thought to operate, involve components internal to the organism. This is especially clear, for example, for mechanisms involving components internal to cells, such as metabolic and/or signaling pathways. However, both at the cellular and organismic levels, several mechanisms operate precisely by crossing the inside/outside boundary. The presence of extracellular molecules can stimulate or repress gene expression, as we know at least since the undercover of operon mechanisms in bacteria. The same kind of phenomenon is present in higher animals, as evidenced in the case of gene dynamics associated with circadian rhythms (), neural plasticity processes (), cell-fate specification in development () or tissue responses to hormone action (). Alongside gene action, normal cellular activities and capacities depend on extracellular states of affairs, as exemplified by the influence of ionic media on the ability of neurons to fire action potentials, or through the exchange processes that occur in respiration.
In the present paper, we would like to delve into what we consider a less explored area, both empirically and conceptually, of mechanistic explanations in neuroscience. Although several important neural mechanisms are internal to the organism, an equally important function of a nervous system is to provide an appropriate coupling between the organism and its immediate surroundings, which we will call 'world' for short. Successful organisms do not live stumbling around but move their bodies in ways that are coherent with their world. They are in a continuous interplay between their internal states and the changes external to them. They respond to challenges at the right time and place, or fairly close to it. We claim that a fruitful way to explain this coupling is by describing the mechanisms that implement it. This raises several interesting issues, such as the question of the body’s (and mechanisms’) boundaries. In addition to an initial conceptual exploration of these topics, we also draft a proposal, based on network analyses of mechanisms, to empirically investigate our ideas.