Second, the idea that there is a “historical law” guiding the development of societies is fanciful. Even if there were some weird sort of pattern which suggested that “liberal” ideas did indeed “win out” in the past, it wouldn’t be anything more than a mere curiosity. It wouldn’t prove anything about liberalism in itself, nor would it say anything about the future. It would just tell us what happened before. To read meaning or predictive power into any pattern in the past is, in fact, about as intellectually respectable as reading tea leaves.
The last sentence is a provocative statement, designed to provoke a response. What is that response? Well, it’s a valid statement if it means “if you see a pattern in the events of the past, then that pattern will be repeated in the future”. That, I agree, is intellectually fanciful. But the broader point, that we cannot read significance in the past, is wrong: the future is produced by the past, and therefore the patterns of the past must have predictive power. As historians, we explain events in the past by finding causes for them. In the same way, historians of the future will explain the events of our own day by reference to causes in our past. Patterns of behaviour originating in our past, continuing through the present into the future will produce that future – at least in part.
But which patterns? That’s the key question. We humans are very good at seeing patterns, and the known past is so vast, so complex, with so many intertwining lives and forces and events, that historians (and non-historians) are able to see an almost infinite number of patterns. In history, we trace historical causation – but historians tend to disagree on what the causes of any one event were. And it is important to recognize there are many causes of even the simplest event, and that we can see “causation” itself in many different ways: we might talk of short- and long-term causes, of preconditions and triggers, and so on.
What this boils down to is this: we are in the present, and we have no way of knowing which of the forces shaping our world will continue to shape the future, and which will be historical dead ends. We know the past will create the future – but we don’t know in what way. So an intelligent European observer in 1785 or 1910 would probably not have predicted that their world was about to crash in ruin. (Or, such an observer in 1715 would not have predicted a generation of relative peace and stability.) There were certainly signs that pointed in that direction – but there were lots of signs, pointing in all sorts of directions. There was, and is, no way to reliably tell the historical signal from the noise.
To look at this from the Roman point of view: explaining the fall of the Roman Republic has been one of the favorite games of historians for centuries. Yet even leaving aside the date we assign to that fall (on which see Harriet Flower’s book Roman Republics, which makes a good case for seeing 88 BC as the end of one particular version of the Republic), it is also legitimate to ask: why didn’t the Republic collapse in 121, with the murder of C. Gracchus? Or in 100 (Saturninus’s killing)? Or in 71, or 67 (occasions when civil war over Pompey’s excessive political demands was nearer than many historians acknowledge)? Indeed, the “Big Question” in republican studies today is not why the Republic collapsed when it did, but rather why it survived for so long. We are able to trace many patterns in the first century pointing towards collapse, but also many pointing towards stability. Questions arising from that are: why did the forces of destruction prove stronger, and would an acute observer at the time (and M. Caelius Rufus always springs to mind as an acute observer in 50) have necessarily predicted that the destructive forces would win out?