Tool-swapping in interdisciplinary research – a case study

By Lindell Bromham

What can we learn from focussing on examples of interdisciplinary research where ideas or techniques from one field are imported to solve problems in another field? This may be in the context of interdisciplinary teams, or it may simply involve borrowing from one field to another by researchers embedded within a particular field. One of the major benefits of interdisciplinary research is the chance to swap tools between fields, to save having to reinvent the wheel.

The fields of evolutionary biology and language evolution have been swapping ideas and tools for over 150 years, so considering the way that ideas have flowed between these fields might provide an interesting case study. The following is extracted from a longer paper examining this area (Bromham 2017).

Language evolution and biological evolution are obviously very similar:

• both can be described by a process of descent with modification leading to hierarchical patterns of similarity between lineages,
• both involve the rise in frequency of individual differences until they form characteristic differences between populations,
• both generate diversity by splitting of lineages which may also suffer extinction.

But language evolution and biological evolution are also obviously very different: where biological evolution acts by the changing frequency of traits inherited by offspring from their parents, language evolution involves changing patterns of intangible sounds, images and concepts, passed between both related and unrelated individuals of any age by any medium, which can be both unconsciously adopted and wilfully changed.

When do the similarities allow us to use the same tools to investigate both language and biological evolution? When do the differences make swapping tools dangerously misleading?

I argue that, as long as we are clear what task we are using the tools for, the fundamental differences between language evolution and biological evolution are in some cases less important than they first appear. But we need to examine each particular case carefully to be sure that the assumptions of the methods are met and that it is doing the job we ask of it.

It is useful to consider three categories of tool swapping:

1. Conceptual tools are analogies drawn between different processes, where the rhetorical device of comparison between similar examples is used to bolster the argument for the reasonableness of one case by reference to the other.
2. Theoretical tools are models that can be used to investigate patterns and processes, using simplified and indirect representations of the systems to explore the influence of different factors.
3. Analytical tools are practical applications, useful pieces of kit that can be profitably employed outside the field they were originally developed in.

Conceptual tools borrowed from linguistics played an under-appreciated role in development of evolutionary biology. For example, Darwin and others used the analogy to language change and divergence to make the acceptance of biological evolution more palatable by comparing it to an observed and widely accepted process.

Theoretical tools borrowed from evolutionary biology, such as mathematical models, have been used in linguistics to model language change over time, even though the processes of change are in many ways very different. As long as the mathematical assumptions are met, the same models can be applied to different situations.

While many analytical tools have moved between fields, such as methods for constructing evolutionary trees and networks, there are some areas where there has been a puzzling lack of transfer, despite the development in biology of practical solutions to problems shared with historical linguistics, such as that of statistical non-independence due to descent.

In addition to these common problems and shared solutions, there are two additional interesting parallels between language change and biological evolution.

One is that the closest that either historical linguistics or evolutionary biology can come to a universal statement is that there are probably no universal patterns of change and every rule has an exception (except perhaps this one).

The other is that extinction is nearly always irreversible, and the loss of either a language or a species represents a tragedy to those who wish to understand evolutionary processes. Each species or language represents the accumulation of unique features and novel combinations of structures, and as such each attests to the relationships between traits, representing the outcome of an evolutionary process under novel circumstances with a unique history, shaped by both chance and directional change. Just as the loss of a language represents the loss of a unique set of words and structures, so the loss of a species is the loss of a unique set of genetic variants that worked together as an integrated whole. Every species or language lost is a thread lost from the global tapestry made over a vast number of lifetimes. The more threads we lose, the less we can understand about the processes and patterns that shaped that diversity.

To find out more:
Bromham, L. (2017) Curiously the same: Swapping tools between linguistics and evolutionary biology. Biology and Philosophy, 32: 855–886

Biography: Lindell Bromham is an evolutionary biologist and professor in the Research School of Biology, Australian National University in Canberra, Australia. Her research spans from the determinants of mutation rate in individual genomes to the generation of global biodiversity over timescales of millions of years. Her experience in interdisciplinary collaborations ranges across projects with artists, philosophers, linguists, mathematicians, medical researchers and veterinarians. She has also developed a way of measuring and comparing the relative interdisciplinarity of research proposals and used it to show that interdisciplinary proposals have consistently lower funding success (see Bromham, L., Dinnage, R. and Hua, X. (2016). Nature, 534: 684-687).