History never repeats

One of my favourite bands of the 80s was Split Enz, out of which Crowded House evolved. And one of their best songs was titled "History never repeats", a sentiment that seems to be fairly widespread. Recently, I started Dawkins' latest book (what is it with established writers on evolution? Gould's brick was immensely in need of editing, and so is The Ancestor's Tale). The first epigram is attributed to Mark Twain:

History does not repeat itself, but it rhymes

Compare this with the somewhat later claim by American philosopher Georges Santayana:

Progress, far from consisting in change, depends on retentiveness. Those who cannot remember the past are condemned to repeat it.

The notion of history as a cyclical, or perhaps iterative, process is ancient, of course. I'll resist the temptation to do the historical review (no matter how badly I want to!), but the notion of an eternal recurrence is, well, recurring in western and non-western thinking.

Oddly, the Twain comment doesn't appear to be from Twain, although I can assure you the Santayana comment is authentic. But why does this view of history seem to persist, and does history, in fact, rhyme?

The notion of history that we have is a mishmash of conflicting ideas. The ancient view that things kept recurring, found in Heine and Nietzsche, is based on the commonsense idea that societies have similarities across time. Toynbee, popular in the early years of the 20th century, especially with science fiction writers, repeated the idea that empires have a predetermined path to tread, and as Rome went, so too will modern societies. This would be, in biological terms, an ontogenetic view of history, a term I choose carefully, because August Comte, the founder of positivism, held that societies literally have a life cycle, that moves from the "theological", to the "metaphysical", to the "positive". He was not the first, but the developmental aspect is pretty clear. There is an inner formative force, an entelechy as it came to be known in biology, that drives cultures and societies through these stages.

Of course, Comte was optimistic. The positive stage was supposed to be the maturation of society, and the cycle here was cross-societal. He didn't think each society underwent cycles. As each society matures to the positive stage, it will settle down into long-term stability. In this respect, Marx is a Comtean positivist. Toynbee's view is more pessimistic - it is inevitable that societies will collapse under their own decadence. They age, and die, in effect, and new ones are born.

Recurrence is based on a macrocosm-microcosm view - as below, so above. This view, which comes to us via the alchemical-mystical tradition of the Renaissance and earlier, holds that large things are just like small things, and so history is itself a kind of organism-lineage, with societies as the organisms. There are strong functional reasons why societies must go through the generation-decay cycle, it is thought.

But historical thinking in the west is also informed by another rather distinct tradition - the Christian, and to a lesser extent Jewish and Islamic, eschatological tradition. Eschatology is the theological subdiscipline that studies the "last days" (eschaton), and if you have last days, and a beginning, then you have an arrow of history. Moreover, there is no need for repetition - the hand of Providence, or the Divine Plan, unfolds as it should, and it may even be there is progress (or the inverse) over time. So the unidirectional view of history is also in our conceptual heritage.

There are many attempts to reconcile eternal recurrence and eschatology. Marx and Engel's dialectic view of history, and its precursors, is unidirectional until the proletariat - the last alienated class that is subjugated to class warfare - is liberated and a classless society arrives. Surely this is the Promised Land, the Kingdom of, well not Heaven but a free society. The cycling of thesis-antithesis-synthesis, of ruling class being overthrown by the next class down, and the new order then setting up a new ruling class, is itself a limited cycling in Marxian terms. Eschatology limits Recurrence.

In modern historiography, the two views are secularised. Recurrence becomes functionalism - institutions exist because similar conditions generate similar institutions. Eschatology becomes a narrative view of paricular histories. Each society and each institution has its own, unique, history. How can we reconcile them?

After the first world war, the naive Enlightenment faith in progress began to dissipate. Researchers began to move away from a positivistic account of social evolution, and stress the relativity, and the particularity, of differing societies. Functional accounts of society became the standard ones, but each functional and structural system was its own kind.

What seems not to have been realised outside the biological community, and not even entirely within it, is that Darwin's notion of what later came to be known as "phylogeny" introduced a novel way of viewing history. Of course, it had been implicit in the view of language evolution first proposed as a tree structure by William Jones, and slightly before Darwin by August Schleicher.

Of course, tree diagrams go back much further than either of these, and the crucial element is time - traditional classifications used a top-down tree structure, while Jones, Schleicher and Darwin introduced differentiation over time instead. What is different about this view of history, is that each branch of the tree is unique, even if there are similarities of function between them. These similarities of function are called "homoplasies", a term coined by E. Ray Lankester, and they obscure rather than elucidate the historical relations between organisms and taxa. But they do, indeed, rhyme.

Darwinian historiography is a late development in the thinking of the west, and it is a hard thing to maintain. Dawkins is trying, despite his tendency to functionalism, to deal with the particular nature of biological history, and the points he makes in Ancestor's Tale are based on this new view, which is best captured, I believe, in the taxonomic philosophy that has come to be known as "cladism". On this view, history is recovered, or at least tested against, patterns of shared derived traits (called synapomorphies in the idosyncratic terminology of the discipline), but you have to eliminate first the convergently derived homoplasies. Once you have a tree, you can work out whether homoplasies have evolved more than once.

Older evolutionist taxonomies would lump together homoplasies on functional (that is adaptationist) grounds. An egregious example was half-seriously suggested by Julian Huxley (grandson of the Bulldog) - Psychozoa (the class of organisms that have reached the "grade of organisation" in which psychological complexity allows cultural evolution). Grade based classifications are fundamentally ahistorical. If meerkats evolved into the grade in twenty million years, they would be members of that grade. Ironically, this is the essay in which the term "clade" (meaning "branch" in Greek) was first coined. Clade based classifications are historical, even though we may not know the exact sequence that generated the clades.

When history rhymes, it is because we have not attended to the differences between particular historical institutions or processes, but the similarities. When we attend to the differences, though, recurrence evaporates. The reason why inferences from someone's or something's being a part of an institution or culture to the shared characters of that culture work is because, basically, they share the ancestry or heritage of that tardition or group. The reason why grade based classifications work is simply because we have pre-specified what it is we want to attend to. In short, the rhymes are in our interests, and the discords or diversity is in the things themselves.

But there'll never be a band as good as Split Enz, again...

Huxley, Julian (1957), "Evolution, cultural and biological", in New bottles for new wine. London: Chatto and Windus, 61-92.