In his book The Integrative Action of the Nervous System (1906), nerve physiologist Charles Scott Sherrington conveyed a holistic conception of nervous action. Where previous accounts had emphasised either that nerves acted entirely according to mechanical principles, or that they were guided by vital forces or materials, Integrative Action developed a non-mechanical but nevertheless physicalist interpretation of their nature.

Cells featured particularly prominently in Sherrington's analysis. The book's opening words declared that 'nowhere does the cell-theory reveal its presence more frequently in the very framework of the argument than at the present time in the study of nervous reactions.'  The same passage went on to suggest that 'the progress of natural knowledge' had enabled biology to pass 'beyond the confines of the study of merely visible form, and... [turn] more and more to the subtler and deeper sciences that are branches of energetics.''

Such emphasis on the physical, microscopic nature of nervous activity was however problematic for many early twentieth century physiologists. Whilst discussions of the physically-derived activities of individual nerves had begun to be be brought to bear in analyses of nervous action, the means by which multiple cells interacted remained far from clear. This essay points to two ways in which, in writing his book, Sherrington drew on traditions that lay outside of the mechanical framwork in which he was initially trained. It shows that on the one hand, Sherrington's appeal to a 'physical' nature allowed him to go beyond the mechanistic conceptions of nerves that prevailed in Britain at this time. On the other hand, it highlights that Sherrington also sought to assimilate 'vitalist' interpretations of nervous action within his more broadly materialist schema. Sherrington's 'integration' was thus in certain respects an intellectual synthesis of two historically antagonistic trends in natural philosophy: the mechanistic and the vitalistic traditions of physiological theorization.

Like other late-nineteenth-century physiologists, for most of his career Sherrington remained firmly within a world in which psychological and physiological researchers shared and contributed to the same conceptual terms and categories of analysis. As Roger Smith has shown, prominent amongst these was 'inhibition'. This term served as a place-holder for debates regarding mental hierarchies. Able to denote both the suppression of nervous response and the control of bodily impulse, inhibition constituted a point of contact between physiological and psychological claims. At stake in discussions of it was not only the possibility that nerve activity might depend on the autonomous actions of psychologically individual cells, but the relation between mind, the brain, and the nervous system more generally. For physiologists, inhibition denoted a negotiation of relations between the latter two of these categories. Was the suppression of certain actions and the manifestation of others due to changes in individual nerves, or to some other cause such as cerebral influence? How, if not by the autonomous linking and de-linking of junctions between themselves, could nerves manifest variable levels of response to stimuli?

Physically-inclined physiologists thereby sought to articulate conceptions of inhibition that a) did not rely on an appeal to vital forces inherent to cells, and b) could nevertheless account for the presumedly psychologically-relevant processes in which particular nerve functions predominated over others. For example, at the turn of the twentieth century, the emergence of osmotic concepts within physical chemistry prompted a re-evaluation by some physiologists of the widely-held mechanistic assumption that the wave-like electrical phenomena produced by nerves was a product of the release of a chemical energy specific to them. Physical chemists working in the 1880s and 1890s had begun to portray electrical potential as a product of the internal dynamics of battery cells. By introducing a semi-permeable membrane between differently-concentrated solutions of liquids, chemists argued, it was possible to model electrochemical phenomena in terms of the production and release of 'osmotic pressure': galvanic current was produced by the migration of ions between differently-concentrated regions of a battery cell. By 1900, physiologists had begun to consider these contentions. Most influentially, Emil Du-Bois Raymond's student Julius Bernstein developed a mathematical model of electricity transmission that did not require any expenditure of cell-specific energy: rather, changes in rates of transmission were due to ionic exchange. For a small minority of physiologists working at the start of the nineteenth century then, bio-electricity could no longer be considered the product of a vital force: rather, it was the result of a calculable alteration in ionic densities between liquids.

With his 1897 appointment as Holt Professor of Physiology at the University of Liverpool, Sherrington was afforded a front-row seat from which he could follow these developments. In 1891, John Smyth Macdonald had been appointed a Holt Fellow at Liverpool under the then-holder of the professorial chair, Francis Gotch. By 1899, he had attained the rank of Senior Lecturer, and had moved away from Gotch's emphasis on relating galvanometric changes to changes in body temperature. Instead, Macdonald began to identify electrical variability with interactions between nerve cells and their environments. Casting the production of current following the lesion of a nerve in terms of an osmotic 'concentration cell', he suggested that the rapidity of transmission of electricity through a nerve was not due to its internal polarization alone, but also to its compartmentalization in relation to its immediate environment. As he reported to the Royal Society (via Sherrington), recent histological staining techniques revealed the presence of ion-carrying potassium granules in the fluids surrounding the so-called 'nodes of Ranvier' (the myelin sheath-free points that could be detected along the lengths of nerves). Whether large or small amounts of these granules could be observed at any one time correlated with the detectability of electrical charge. Macdonald suggested that nervous current was therefore produced not within a single, undifferentiated cell, but by interaction between a series of 'relays placed at every point of the nerve to ensure the continuous propagation of the excited state.' Significantly, these relays presented an explanation for the both the variability and the unidirectionality of nerve conductivity:

the forward movement of the negative charge is ensured, or rather its backward transmission is prevented, by the pursuing positive charge... the rise and fall of osmotic pressure respectively lag a definite time behind the appearance of the causes producing them... Arranged in these terms the wave of the nervous impulses can be described as a double oscillation in the value of osmotic pressure, the front a rise, the trough a fall.

Like his Cambridge contemporary William Bate Hardy, Macdonald doubted the existence of an autonomously-acting protoplasmic network extending outwards from nerve cells. He proposed in its place an explanation of the variability of nerve cell conductivity which appealed to the changing proportions of ion-carrying compounds in the fluids within and surrounding cells. Inhibition of individual nerve cells was thus due to 'a reversible change during which electrolytes are set free into a state of simple solution, and are then recovered from this state back into their original condition.' The vibrations detected by protoplasm theorists were not the products of organic fibrils connecting inner life with its external conditions, but changes in the ionic concentrations of fluids behaving according to mathematical laws of osmosis.

Perhaps unsurprisingly given his physical physiological commitments, Sherrington invested significantly in Macdonald's research. This was to a great extent due to the possibilities that he saw in it regarding a more general explanation of physiological inhibition. After a consideration of Gaskell and Hering's metabolic propositions, Sherrington devoted nearly three pages of Integrative Action to Macdonald's conclusions. Summarizing the latter's research as leading to the proposition that 'inhibition is the condition in which the possibilities of free motion are most reduced', he suggested that Macdonald's views were 'fertile in suggestion for future experiment.'

Such enthusiasm reflected a new emphasis within physiology on the study of nerve junctions themselves as sites at which nerve conduction might be interrupted. In his Textbook of 1900, Edward Sharpey-Schäfer thus suggested that impulses were 'momentarily arrested at these places of contact of nerve cells with one another.' Sherrington, characterizing such arrests as manifestations of a 'neurone threshold', identified it unequivocally with a non-contiguous, 'synaptic' transmission:

at each synapse a small quantity of energy... acts as a releasing force to a fresh store of energy not along a homogeneous train of conducting material, as in a nerve fibre... but across a barrier which whether lower or higher is always to some extent a barrier.

This insistence on the functional significance of barriers between cells marks a critical change of emphasis within physiological considerations of inhibitory mechanisms. Where explanations of variations in the rate of electrical transmission along nerves had centred on cells and the substances that composed them, Sherrington and his colleagues characterized such variability as dependent on changes in the thresholds at which transmission between cells could be effected. Neither cell bodies nor protoplasm were the most physiologically significant conditions for variation in the transmission of nervous impulses: rather, 'synapses' were.

It would be a mistake however to suggest that Sherrington relied on strictly physicalist conclusions regarding nervous action to establish his contentions regarding inhibitory activity. Significantly, in the passage of Integrative Action immediately following that recommending Macdonald's research, Sherrington moved on to discuss a somewhat more complex aspect of the variation of nervous function. This section, which he disassociated from inhibition proper, concerned the effects of stimulation of more than one nerve at the same time.

It was well known that if particular areas of an experimental animal's skin were stimulated, one response could be elicited for a certain time, following which an entirely different response might take over. Stimulation of a set of points or areas of skin could elicit radically different actions depending on their situation and longevity. Sherrington had sought to make this area of investigation his own. In a long series of notes presented to the Royal Society between 1893 and 1909, he set out to define the nature of the so-called 'knee jerk' reaction, and with it what he would call 'antagonistic muscle action.' As a reaction that was dependent on the simultaneous action of two muscles working in opposite directions, this topic presented a more complex problem than those involving excitation of a single nervous pathway. The variability of antagonistic reactions, Sherrington found, was dependent on the initial posture of the animal, the extent to which stimulation had previously been applied, and the strength of the stimulus. Particularly notable were situations in which two reactions would alternate between one another. In Integrative Action, such dynamic alterations constituted a significant stepping-stone between the explanation of the inhibition of simple nervous reactions, and that of reflex activities more generally.

As set forth in 1906, however, Sherrington's explanation of alternating 'antagonistic' reactions appealed not primarily to his own studies, but to the research of the physiologist and psychologist William McDougall. As with Macdonald, Sherrington knew McDougall personally: one of McDougall's first actions on his 1904 appointment to the Wilde Readership of Mental Philosophy at Oxford was to invite Sherrington to lecture there. McDougall's explanation of reflex action was heavily indebted to Sherrington's studies of nervous action. Moreover, it identified the critical variable in the alteration of nervous response in functionally significant barriers between cells. Though previous studies had characterized inhibition as an arrest of assimilatory or (in Gaskell's terminology) 'katabolic' activity, McDougall noted, the mechanism for this arrest was far from obvious. Noting that 'the inhibition of a mental process is always the result of the setting in of some other mental process', McDougall ascribed the inhibition of all nervous processes to the predomination of others. There was no specific inhibitory force working against the activity-generating forces of cells: rather, inhibition occurred at point of diminution in the number of nervous paths. Thus, given two 'antagonistic' nerve arcs entering a synapse, the manifestation of one of these would prevent the other from gaining expression. Similarly, when an initially-predominating reflex diminished in strength, the nerve arc that had not yet found expression would take over. Here then was a conclusion that appeared complementary to Macdonald's studies. As in Macdonald, changes in nervous response were to be identified not with the activities of individual nerves, but with changing conditions at the sites of interaction between them. Synapses were sites at which peripheral nerves competed for access to what Sherrington termed a 'final common path' that would register their stimulation more widely through the system.

Crucially, McDougall had arrived at his conclusions not primarily through experimental or anatomical study of nerves, but that of his own sensory impressions. After a period on the Cambridge University-led Torres' Strait expedition and a short-lived career in medicine, he had sought out the Göttingen psychologist Georg Elias Müller, whom he later described as 'then the leading exponent of the exact laboratory methods in psychology.' This training had brought him into contact with the few British psychologists who had begun to engage with German experimental psychology, including W.H.R. Rivers, and James Sully. In an effort to promote the experimental study of mind in Britain, Sully had in 1897 acquired a set of laboratory equipment from Hugo Münsterberg that the latter had created for his own psychological laboratory in Freiburg. Without any expertise in experimental practice of his own, Sully persuaded McDougall to return to Britain to take up research at University College London. This in 1903 led to his appointment there as Reader in Experimental Psychology. McDougall had thereby spent the first years of the twentieth century living in a small house on the Surrey Downs, engaging with a set of tools developed within the then-emerging German tradition of psychological research, and adapting his conclusions regarding these to the British context.

Furthermore, McDougall conveyed his findings in terms that went directly against the physicalist tendencies of late nineteenth century British physiological physiology. He began his career by rejecting the widely-held proposition amongst British physiological psychologists that mental faculties could be identified with one or another anatomical part of the brain or nervous system. Instead, he proposed, awareness was ‘immediately determined' by 'neural processes' as a whole. Thus an increase in nervous excitability could be attributed to a 'diminution of the resistance offered by that delicate and complex inter-cellular substance which, as it seems to me, we have to regard as the seat of the psycho-physical processes.' This substance, according to McDougall, was a vital fluid which he termed 'neurin.' On the stimulation of a sense, its associated nerves ‘[produce] neurin far more rapidly than it can escape by leakage’ into adjoining cells. Build-ups of neurin overcame the resistance of synaptic barriers, causing chains of like reactions through the whole system.

In his vision studies, McDougall deployed this physiological schema to account for a set of illusion-generating experiments that he encountered amongst Münsterberg’s laboratory equipment. The most significant of these for McDougall concerned the phenomena of variations in spatial attention as it related to vision. Following Hering, Münsterberg had proposed that the alternation of attention between different sense-organs could be explained in terms of a pair of generally-acting antagonistic forces within the body: the motor functions of the eyes, for example, strained against one another to apprehend their surroundings. He had thus developed a series of experiments that showed ways in which visual effects could predominate first in one eye, and then in the other. McDougall re-interpreted Münsterberg's conclusions in terms of a fluidic 'X-substance' within the retina. The transition of attention from one eye to another was not due to the temporary predominance of assimilatory over dissimilatory forces, as Hering and Münsterberg had claimed, but rather to the interaction between the stimulatory effects of this substance as it was conveyed via the neural fluid. The alternation of images between the eyes could be explained by the interaction between differently-originating streams of this fluid competing for synaptic influence: 'owing to the mutual antagonism of the... [neural] systems, that of one inhibits that of the other'. Moreover, if an area of the retina was exposed to light for a long time, the X-substance in that area would be used up, allowing another stream of fluid to predominate. McDougall thereby characterized spatial interactions between the visual fields in terms of the inhibitory activity or 'drainage' of vital fluids that coursed from the retina through the nervous system. Visual sensation was cast in terms of autonomously-acting vital substances of the same order as that which underpinned Hering's conclusions, and to which Bergson would also appeal.

Perhaps unsurprisingly given the above, Sherrington's relationship with McDougall appears to have caused consternation within Britain's physically-oriented community of experimental physiologists. For example, Sherrington's predecessor in the Waynflete chair, John Scott Burdon-Sanderson, wrote to the former shortly before his 1904 appearance, noting that he had 'just become aware that you are to deliver a lecture here' at the behest of McDougall, that he could not 'be sure whether it is to be on Physiology or Experimental Psychology', and requesting that Sherrington attend luncheon with him after the lecture. He further added, with something seemingly approaching sarcasm, 'if it would suit you... to dine with us tout niveux [sic].' Sherrington had spent the previous two decades building up a reputation as one of the most innovative physiologists in Britain, and identified Oxford as a place in which he might further his own physiological goals: he had even unsuccessfully sought appointment there in 1895. Burdon-Sanderson could thereby be sure that his expression of uncertainty regarding Sherrington's disciplinary commitments would be well attended to. Most likely mindful of the polite warnings of his physiological colleagues, Sherrington was careful to avoid uncritical reliance on conclusions arrived at by the investigative means that McDougall employed.

Despite his enthusiasm regarding McDougall's explanation of 'reciprocal innervation', then, Sherrington's characterization of it limited its explanatory scope to a particular aspect of a broader, physical, phenomenon. Sherrington admitted that McDougall's conclusions presented 'an explanation for the transition from one antagonistic reflex to another'. As a framework for inhibition, however, they tended

to sever... central nervous inhibition – of which I regard reciprocal innervation of antagonistic muscles as but one widely spread case – from other forms.... It appears to me unlikely that in their essential nature all forms of inhibition can be anything but one and the same process.

Other forms of inhibition, a uniformly-caused function of nervous action, could not be explained by McDougall's scheme. McDougall's psychological insights were re-cast in Integrative Action as contributions to the experimental study of physical nervous function.