Evolution: Trees and Braids.

One of the most powerful visual tools for thinking about evolution is the tree. When scientists want to present evolution of any kind they typically show this kind of diagram which branches as it travels  from bottom of the picture upwards, becoming every more diverse as new lines split from old. The upwards motion itself also invokes metaphors, particularly "up is good" and its corollary "down is bad". We forget that every living thing presently alive, from bacteria to blue whale, is the product of 3.5 billion years of evolution and thus equally evolved.

The tree diagram shows us for example that humans and chimpanzees had a common ancestor some 5 million years ago, and indeed that all life seems to have evolved from a single kind of organism. We see the same paradigm in physics and diagrams of the evolution of the universe and in the study of ancient texts. Diversity of similar objects in the present seems to automatically imply a common ancestor.

Buddhists use to paradigm to point to the common origins of Buddhism. However in my last essay I cited Paul Harrison's comments on the Vajracchedika: "...to put it in a nutshell, the idea that the wording of any Mahāyāna sūtra can be restored to some original and perfect state by text-critical processes must be abandoned: all lines do not converge back on a single point." (Harrison 240, emphasis added) 

In this essay I will try to show that the tree diagram inevitably falsifies evolution and other complex developmental processes. The project of making lines converge is not always able to account for the complexity of reality. I will also propose another, better, metaphor for conceptualising and visualising these processes. In writing this I also have in mind a discussion on Sujato's blog about how we interpret and make conjectures about the origins of Buddhism based on the textual evidence which dates from some centuries after the putative origin.


Horizontal Links

Let's begin with bacteria. The standard view of bacteria is that like other forms of life they evolved into thousands of species which can be classified in the standard taxonomies, using the standard Latin nomenclature: Genus species. For example Bifidobacterium longum occurs in the gut of infants and plays several roles including breaking down the complex sugars in milk to help the infant digest them. While Streptococcus pneumoniae is a very different bacteria that colonises the lungs and other tissues and causes pneumonia and also meningitis. These two bacteria have very different habits. And yet Lynn Margulis argued that bacteria have no species because they can all shared genetic material.

Biology blogger, Julius Csotonyi, has called them "Plagiarizing Wizards", Csotonyi's account of "horiziontal gene transfer" is amusing and informative at the same time. Any bacteria can share genetic material with any other bacteria and some viruses (which Lynn Margulis characterised as stripped down bacteria). It is partly this ability that geneticists employ when the insert or remove genes from organisms. This assimilation of genetic material changes the organism. We might say that in assimilating foreign genes they had become a new species, just like that.

Furthermore, Csotonyi says "Even more amazingly, there is evidence that under stressful conditions (e.g. heavy metal-polluted waterways), the rate of horizontal gene transfer between bacteria increases, as if stress induces a more urgent swapping of genetic ideas for a solution."

Graphical representation of horizontal gene transfer. The branched tree-like structure represents the evolutionary lineage (geneological tree) of representatives of earth's major types of life forms. Sometimes genes can be transferred (horizontal gene transfer) between otherwise distantly related species. This is illustrated by bridges forming between branches of the tree, where genes 'jump' from one lineage to another. 

Image: Barth F. Smets, Ph.D.

via Evolutionary Routes

Bacteria always live in communities of many 'species' or as we ought to say 'varieties'. And these varieties swap genes. Each gene codes for a protein which performs a specific task. It may be a structural element, or an external marker used for communication, or very often it will be an enzyme which facilitates and/or catalyses a particular kind of chemical reaction. In the heavy metal example, a protein might chelate a heavy metal atom - i.e. warp it in an organic molecule that effective seals it off from the chemical environment surrounding it, rendering it inert. Chelation is the first line medical treatment for heavy metal poisoning. And this ability which one bacteria has, can rapidly spread through a whole population of bacteria under ideal conditions. For example, this is how bacteria can acquire immunity to antibiotics. It only takes one bacteria to express a gene that produces a protein that neutralises the antibiotic agent. That bacteria survives in a situation of drastically reduced competition for resources and thus breeds rapidly, but also passes on the gene that makes it successful.

Thus the tree structure cannot describe the process by which the current variety in the population of bacteria occurs. The diagram looks more like the image on the left. The technical term is a reticulated network, but below I will propose a metaphor drawn for nature for it.

In the diagram above, higher level structures such as mitochondria and plasmids are also shared between varieties. Further up the taxonomic ladder we strike the phenomenon of hybridisation. We are probably all familiar with the popular, and useful, definition of a species that says that two organisms are different species if they cannot breed and produce viable offspring. Thus a horse and a donkey can produce offspring, mules, but they are sterile and we consider them different species. We also see lion and tiger hybrids in captivity producing sterile, so called, ligers. Wolves and dogs on the other hand produce viable offspring, and as a result the domestic dog has been reclassified as a sub-species of wolf. Hybridisation is far more common than has previously been suspected. Reporting on an article in Nature the National Geographic News said "on average, 10 percent of animal species and 25 percent of plant species are now known to hybridize." Of course most times these inter-species matings result in infertile offspring. But not always. When the offspring are fertile then a new species is born. Off course the chances of successful hybridisation are low, but they seem to be considerably higher than the likelihood of a beneficial mutation in a single gene, let alone the accumulation of such mutations. 

Clearly any hybridisation event that is viable produces a cross-link in the "tree". Certainly in the plant kingdom where this is going on in 1 in 4 species the result is going to be a highly cross-linked reticulated network, rather than a tree. For animals less so, but the cross-linking is going to outweigh the splitting supposed to be caused by beneficial mutations by orders of magnitude. 

Family tree of the four groups of early humans living
in Eurasia 50,000 years ago and the gene flow
between the groups due to interbreeding.
Image credit: Kay Prüfer et al.

One of the fascinating genetics/evolution stories of recent years was the discovery in 2013 that modern humans in Eurasia appear to have Neanderthal DNA indicating that our two species Homo sapiens and Homo Neanderthalensis interbred to some extent (see Discover) before Neanderthals went extinct. We Eurasians, then, are hybrids of H. sapiens and H. Neanderthalensis. There was also apparent interbreeding between other late varieties of modern humans such as Homo Denisova and Floriensis (see also here and here and this). It's getting more and more difficult to maintain the the genus Homo can be represented by a simple tree. The hominid family tree is highly cross-linked and in fact resembles a reticulated network. That said all present day humans are considered to be the same species and subspecies, i.e. Homo sapiens sapiens. 


A Braided Skein

I've often commented on the way that Buddhism has hybridised with the cultures that surround it. If the Iranian Origin thesis is correct then the Śākyas arrived in the central Ganges plain at a time (ca. 850 BCE) when it was extremely culturally diverse. First wave Indo-Aryan speakers had already begun to dominate over local speakers of Tibeto-Burman, Austroasiatic, and possibly Dravidian languages, but the latter were still present. Second wave Indo-Aryans (the Vedic speaking Brahmins) were already moving east and starting to influence and be influenced by the cultures there. Not only this but there is evidence of considerable genetic variation in India also (e.g. Ethnic India; DNA Testing). I've described the emergence of Buddhism, Jainism and Ājīvakism (and to some extent Upaniṣadic Brahmanism) as the culmination of a process of assimilation and synthesis of elements of the cultures associated with these various languages. So by the time Buddhist texts are composed we see influence from Brahmins in the form of gods such as Indra, Brahmā, and (probably) from Austroasiatic in the form of local spirits associated with water and/or trees such as yakṣas and nāgas for example. In the Buddhist doctrine of karma we can see influences from Brahmanism and Jainism along with remnants of Zoroastrianism (hope to get this conjecture published soon). And we know that such hybridisation continues. Buddhists continue to borrow elements from other cultures and religions down to the present. 

Rakaia River on its way
to the sea.

Even if the Iranian Origin thesis is wrong, the course of the development of Buddhism is clearly not a simple tree structure resulting from internal splits. generating the traditionally names sects. No doubt there were internal splits, but there was a great deal of hybridisation as well. The tree image is inadequate to describe such development. Taking my cue from Indian use of river metaphors, I have envisaged the development like the course of a braided river system. There is an overall flow in one direction, but the flow constantly branches and recombines across a broad bed. There are no straight lines. At times there seems to be a "main-stream" and at other times no one stream dominates, the patterns of branches and convergences is constantly changing.

And of course even if the Śākyas did originate in Iran, things were by no means simple there and then. The influence of Egyptian religion, particularly in the matter of eschatology (or afterlife) for example is quite obvious. Witzel's method of comparative mythology, in his book Origins of the World's Mythologies, purports to trace the roots of our story telling to Africa ca. 65,000 years ago as the first (successful) migrations of modern humans into the rest of the world commenced. And so on back and back until we can no longer determine any source. 

Metaphors and Schemas

One of the basic schemas identified by George Lakoff by which we organise and conceptualise our experience is the origin-path-destination schema. The tree diagram is a variation on this basic schema adding binary divisions and multiple destinations to a single point of origin. And it seems natural for us to organise information along these lines because the schema is one of the fundamental patterns we use to build conceptual metaphors. The origin-path-destination schema is something like a Kantian a priori. But in this case the metaphor does not quite fit reality. We need to invoke another schema to better fit our experience. I suggest that in terms of fundamental human experience the schema that best fits is the community made up of a number of inter-marrying families. As time goes on certain characteristics tend to be retained in families over generations, but at the same time characteristics morph and change because of intermarriage. Likewise the community may remain relatively stable as an entity over many generations despite continual changes in personal due only to birth and death.

The concept of common origins, of seeking for common origins certainly has power at times and it certainly has a powerful grip on our imaginations. But it always over-simplifies the origin. Whether we are dealing with evolution as a whole, the human species, or the products of our culture like the Buddhist religion, or the texts produced by our Buddhist ancestors, there is extremely unlikely to be a simply origin. We Buddhists in particular want to trace everything back to one man. But that one man was just as much a product of his conditioning as any human being. We all have to learn the language and the ways of our family, community and nation. In a multicultural environment like 500 BCE Ganges Plains, or 21st century UK, we also have to engage with differences. Looking backwards there are always continuities and discontinuities and hybridisations. Whoever the Buddha was, he was a member of a family and a community that shaped him just as we were shaped by our families and our communities. Indeed one of the implications of the Iranian Origin thesis is that we should place more emphasis on the culture of the Śākyas as a community as the source of Buddhist beliefs, especially regards morality, and less on any one individual. This might explain why a name had to be invented for the founder at a later date. It is interesting that Buddhists were often known as Śākyans in ancient India - the early medieval Mīmāṁsā thinker Kumārila refers to Buddhists as 'the Śākyas'.

Even if we can point to a single founder, he himself was the product of complex processes. When we see the Buddha as a like a spring (a origin-path metaphor) we actually falsify what we know about every human being - even the most remarkable people are shaped by their environment, by teachers, by family, by history. In fact since he wrote nothing, it was the Buddha's followers who shaped our views of what Buddhism is. What they remembered, what they emphasised, and what chance allowed of that subset to survive is at least as influential as the Buddha himself presuming he existed.

Buddhism is the product of complex historical and cultural processes - a braid rather than a tree.

~~oOo~~

Updates

31 Dec 2013: Seems my use of the word "braid" was on target: Viewpoint: Human evolution, from tree to braid by Professor Clive Finlayson. "Some time ago we replaced a linear view of our evolution by one represented by a branching tree. It is now time to replace it with that of an interwoven plexus of genetic lineages that branch out and fuse once again with the passage of time."

9 Jan 2014. I've also just watched this, quite high level, presentation featuring Lynn Margulis who is also very critical of the tree metaphor for describing evolution - indeed she is a major inspiration for this outlook. 

28 Mar 2014. See also my essay Extending the River Metaphor for Evolution.  

22 Apr 2015. Sweet potato naturally 'genetically modified' Eurek Alert. "Sweet potatoes from all over the world naturally contain genes from the bacterium Agrobacterium."  These genes have been transferred to the genome of the sweet potato and are passed on when it reproduces. This more evidence that a simple linear tree does not describe the processes that go on in evolution. 

29 Apr 2015. Paul Heggarty, Warren Maguire and April McMahon. (2010) Splits or waves? Trees or webs? How divergence measures and network analysis can unravel language histories. Philosophical Transactions of the Royal Society. B. 365. doi: 10.1098/rstb.2010.0099. 

Abstract: Linguists have traditionally represented patterns of divergence within a language family in terms of either a ‘splits’ model, corresponding to a branching family tree structure, or the wave model, resulting in a (dialect) continuum. Recent phylogenetic analyses, however, have tended to assume the former as a viable idealization also for the latter. But the contrast matters, for it typically reflects different processes in the real world: speaker populations either separated by migrations, or expanding over continuous territory. Since history often leaves a complex of both patterns within the same language family, ideally we need a single model to capture both, and tease apart the respective contributions of each. The ‘network’ type of phylogenetic method offers this, so we review recent applications to language data. Most have used lexical data, encoded as binary or multi-state characters. We look instead at continuous distance measures of divergence in phonetics. Our output networks combine branch and continuum-like signals in ways that correspond well to known histories (illustrated for Germanic, and particularly English). We thus challenge the traditional insistence on shared innovations, setting out a new, principled explanation for why complex language histories can emerge correctly from distance measures, despite shared retentions and parallel innovations.

Also http://rstb.royalsocietypublishing.org/content/365/1559/3829