Home > Language, Rivers > Language flows

Language flows

Rivers are good to think with. They are useful not only as metaphors, but also as exemplars of flow. Theories about a wide variety of past phenomena can to some extent be modelled on the physical characteristics of river forms and flowing materials.

Actually, we already use rivers in our thinking, perhaps more than we realise. Take historical linguistics as an example. A central assumption of the discipline has long been the idea that groups of languages have ‘sources’ – or ancestral languages from which more recent and contemporary languages are descended . As Renfrew explains, it is argued that

“there were a number of related languages (Greek, Latin, Sanskrit) and sub-families (Celtic, Germanic) which must be ‘sprung from some common source’, thereby implying both the notion of an earlier proto-language (such as Proto-Indo-European) and an original homeland. Such concepts are now basic to the discipline of historical linguistics” (Renfrew 2000, 10).

The point is that if you map the relationships between developing languages according to such ideas, the result is something very much like part of a river system. Diverging paths of flow are envisaged to run in a branching pattern that resembles configurations of river channels on deltas and alluvial fans, sometimes called distributary networks.

On this model, similarities between disparate languages can be explained in terms of deriving from a hypothetical proto-language, in the same way as different branches of a river (especially in its lower reaches) may have diverged from a single course further upstream. It is tempting to link such divergences of language flows with movements of peoples and the cultural diffusion of ideas and techniques, such as the spread of farming across Neolithic Europe from the Fertile Crescent area – and the reader is referred to Colin Renfrew’s critical discussions on these matters (Renfrew 1987, 2000).

But as Renfrew points out, such diverging models do not seem nearly so applicable to deeper prehistory. Before the advent of farming and village settlement, groups of hunter-gatherers were small and geographically remote from each other. It is hard to imagine such scattered groups speaking a common language: on the contrary, on the model of divergence outlined above it seems likely that there was a multiplicity of different languages. Only with the coalescence of people into larger communities during the Neolithic would these languages have come together. Not so much a divergence, then, but rather a convergence of disparate languages. Instead of an ever increasing number of languages, the number of different languages would decrease as they merged into ever larger flows. This idea can be expressed very simply by comparing it to those parts of river systems where smaller streams or tributaries feed into larger ones in converging branching structures called tributary networks.

At first sight the diverging and converging models might seem to be diametrically opposed to each other, and indeed the two forms of explanation have clashed in the context of historical linguistics debate. Speaking of recent developments in the discipline, Renfrew notes the increasing prominence given to convergence models and the tension between these and more established divergence models (Renfrew 2000, 13).

Rivers are useful here in showing us how apparently contradictory models of divergence and convergence can be integrated together. River systems display the characteristics of both distributary and tributary branching networks. The type of network which prevails on any given stretch of river depends on local geomorphological conditions and the ‘stage’ of the river on its journey to the sea. Thus many rivers configure themselves in a primarily tributary pattern in their upper and middle reaches. This changes to a distributary pattern in lower reaches where rivers drop much of their sediment load, and break up into numerous channels to flow around it into the sea.

It could be argued that language flow works something like this too, adopting the overall form of a tributary network or distributary network at different times according to circumstance. As in the case of rivers, both forms of network may co-exist, with neither being mutually exclusive of the other. But so far we have been thinking of distributary (diverging) and tributary (converging) networks manifesting at different times in the flow of a language through time. Now I want to turn to another sense in which diverging and converging branching networks can manifest simultaneously. Rivers can show us how.

We tend to understand rivers mainly in terms of natural processes. But as I argue in my new book, Fluid Pasts (Edgeworth 2011) they are actually complex entanglements of natural and cultural forces, and need to be re-mapped accordingly. In their middle reaches there is not just the classic tributary pattern of river channels to take into account. There is also the host of partly artificial distributary channels designed to take flow from the river to serve a whole variety of functions – head-races of mills, town boundary ditches, inlets to industrial cooling plants, irrigation ditches, drainage cuts, navigation channels etc – many of which split up into smaller channels before being returned to the river further downstream, often in the form of drains. The crucial point here is that these distributary channels are important parts of river systems, fully integrated into the tributary network, even if usually left out of maps of rivers. Any meaningful discussion about flow regime has to take these into account.

The sketch shows the middle reaches of a typical tributary river network with distributary network added. Actually, the two networks are really just a single network, in the context of which most stretches of river are at once tributary and distributary streams. Most rivers in their middle reaches actually look much more like this than they do the classic tributary pattern so often illustrated in geography textbooks.

Rivers thus provide us with useful working models for understanding how diverging and converging flows can operate simultaneously in the same places and the same times and as part of the same processes. Can this help our understanding of the development of language through time? I believe it can. Thinking of language flow along the lines suggested by rivers, it is not contradictory to hold that languages converge and diverge at the same time. Models that emphasise divergence of related languages from a common source or proto-language can be quite compatible with models that emphasise convergence.

Not just a metaphor

Comparing language and rivers may seem counter-intuitive at first, but actually it makes a lot of sense. It is not just a metaphor. Rivers and languages are very similar (in the sense of being forms of flow) and more closely entwined in a material sense than is usually acknowledged. Because rivers afforded both avenues of movement and obstacles to movement for people, many languages effectively travelled along, were grouped around or divided by rivers. Evolving river systems over the last few thousand years have been inextricably entangled with evolving languages. Take a look at this map of the Amazon and Orinoco river basins showing principal indigenous language families of the region. Although the map gives a greatly simplified picture, one can see how the branching network of languages comes to take on roughly the same shape as the actual river basin, with language flows and river flows woven together, unfolding on shared tangled trajectories. But that’s another story…


Edgeworth, M. 2011. Fluid pasts: archaeology of flow. London, Bloomsbury Academic.

Renfrew, C. 1987. Archaeology and language: the puzzle of Indo-European origins. London, Pimlico.

Renfrew, C, 2000. ‘At the edge of knowability: towards a prehistory of languages’ Cambridge Archaeological Journal 10 (1), 7-34.

  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: