Veronica Strang has written an excellent account of the fluid relationships between humans and the non-human materiality of water. The very idea of anthropology of water challenges our commonly held assumptions about water being just a material or economic resource (or, conversely, about anthropology being just the study of people). It holds that water is socially, politically and ritually constituted, while acknowledging at the same time that water also has an existence outside the human domain. There is much to agree with and to be inspired by in the paper, which is especially strong on ‘fluidity and consistency at every level of human–non-human engagement’ (p. 133). By virtue of the emphasis on that specific aspect, however, there are some aspects which are underemphasized, and it is these which I focus on in my comments.

I refer mainly to rivers, as this has been the focus of my archaeological research. Rivers are interesting because they not only comprise materials, in the form of water or mud or biological inhabitants of riverine environments. They also are characterized by an energy that is derived partly from the sun and partly the gravitational pull of the Earth, which together power the hydrological cycle (with tidal stretches of rivers also subject to the influence of lunar gravitation). Put your hand into a river, or immerse any part of your body, and you can feel these cosmic and planetary forces at work in the form of the flow or current of water. Put a dam or other artificial structure across the river or modify the channel in some way and you actively interpose human projects and designs onto the supposedly natural cycle. Put a waterwheel or turbine or other kind of energy-utilizing device in and you can turn the flowing energy towards some particular task or end. It is more than merely the materiality of the water itself that we are dealing with here. It is more than human agency, too, or the relationships between human and material, however fluid these may be. It is also the flow or energy of the river itself.

For this reason I find the discussion on the agency of water, while useful in many ways, somewhat constrained in others. Established arguments about the material agency of things are applied as though water is little different from other, more solid, kinds of material. Yet water in general, and rivers as flowing energies in particular, have so much more to offer to a theory of material agency. The agency of a river is quite different from that of objects that reside in a museum cabinet, though I do not deny that these can have power of a kind too. To experience the difference at first hand, just push a boat out into a river until it gets caught by the current. The river itself moves the boat in a particular direction, sometimes (unless the eddying force is actively countered) spinning it around. The agency it exerts or force it applies is palpable – something that has to be responded to in order for the person in the boat to regain partial control of the situation. Anyone engaging with rivers, on whatever scale, finds themselves caught up in this physical interaction with river forces. And for all the fluidity of the interactive situation, there is actually something dialectical about it – a sort of ‘push and pull’, or ‘action and counteraction’ – which unfolds through time (Edgeworth Reference Edgeworth2011).

In looking at the archaeological evidence of human–river interactions in the past, a typical scenario is as follows. People make some material intervention in rivers, but this leads to an unexpected river response, which in turn requires some human reaction, which results in further unexpected river responses, and so on. Both human and non-human participants get drawn into a dynamic entanglement or enmeshment with each other – almost a kind of wrestle.

A recent historical and contemporary example is the ongoing struggle between the US Army Corps of Engineers and the Mississippi River to stop it flowing into the Atchafalaya River, thus preventing a change of course and the consequent shifting of the delta, with all the economic and political shifts such a scenario would entail. The engagement perhaps began in earnest with the hand-digging of a small channel in 1831 to provide a short cut for steamboats, and the use of explosives to break a logjam on the Atchafalaya River in 1839. But the unanticipated side effect was that it presented the Mississippi with a shorter and steeper route to the sea, pulling more and more of its waters out of the established channel. Each river-control structure built to prevent the change of course brings about further river responses, requiring further or strengthened river-control structures to counter them. The unfolding human–river interaction is well documented by McPhee (Reference McPhee1989).

A more archaeological example is provided by the discovery of the material remains of (a sequence of) three medieval bridges next to the River Trent at Hemington in Leicestershire. Bridges present both openings and obstacles to river flow, sometimes creating vortices which scour out holes beneath bridge piers, undermining them. Stratigraphic evidence indicated that such scour holes had led to weakening of parts of the structures, contributing to their eventual destruction during episodes of flooding. Each successive bridge used different construction techniques in an attempt to respond and adapt to river conditions, but the river was changing too, partly in adaptation to the bridge structures themselves. Attempts to counteract the force of water, sometimes by presenting material resistance to it and sometimes by the alternative strategy of presenting as little resistance as possible, are evident in the design of bridges. The whole sequence of three bridges and related evidence is a material testament to the dynamic and temporally unfolding character of human–river interaction (Ripper and Cooper Reference Ripper and Cooper2009).

I do not mean to suggest that all human–river interactions should be configured as a battle between oppositional forces. That would be a caricature, for people work with flow as well as against it. But something important needs to be added to the analysis of relationships between the human and the non-human. Partly it is the temporal dimension – and this is where archaeological and historical approaches to river flow can augment the anthropological one – but it is not just that. It is also the sheer power of water, which goes far beyond just material resistance to human agency. This is a power that, at its most destructive and dangerous, can ravage, inundate, overcome and submerge cultural forms. The torrential flooding of the River Arno that swept through Florence in 1966 is the other aspect of ‘the mesmerizing shimmer of water surfaces, the excitement of waterfalls’ (p. 136) just as the tsunami that hit the east coast of Japan in 2011 is the darker side of ‘the calming rhythm of waves’ (p. 136). Even when not in such destructive mode, flowing water still actively and physically challenges human projects and intentions. Friction and turbulence, as well as fluidity and consistency, can result. Rivers represent a force to be reckoned with.

Another way of putting it is that, for all that rivers are partly shaped by humans, they retain something wild and non-human about them. The dangerous, unpredictable aspect of river water is still there when using it as a metaphor or model for reconceptualizing human–non-human relationships. Strang rightly points out that ‘thinking with water’ can help reconnect otherwize polarized positions and perspectives, especially with regard to closing the gap between concepts of culture and nature, human and non-human. Following from the discussion above, and looking at the author's own power to challenge viewpoints through her writing, it is worth adding that thinking with water also has the capacity to subvert established categories, to undermine long-held assumptions, to flow around and over static structures of classification and analysis, to break out of old and established channels – and thus to carve new paths of flow, new ways of thinking.