More examples of unreliable narratives, radiocarbon dating
Near the border between Scotland and England lies a 73 mile wall. Conventional chronology tells us it was built by Emperor Hadrian in 117 AD. Given the sheer size of the thing and that it’s a key feature of history taught to British schoolchildren, how much evidence exists to support these apparently basic facts?
Surprisingly, there are just two pieces of evidence in support. The first is the testimony of a book called the Augustan History. Beyond the usual chain-of-custody problems documented in the first part of this series, this work has another unusual issue — historians are in agreement that much of it is fiction and the authors probably didn’t exist. Wikipedia says:
Major problems include the nature of the sources it used, and how much of the content is pure fiction. For instance, the collection contains in all about 150 alleged documents, including 68 letters, 60 speeches and proposals to the people or the senate, and 20 senatorial decrees and acclamations. Virtually all of these are now considered to be fraudulent.
This book is the primary source of information on Emperor Hadrian. Whilst other classical authors mention him briefly, those texts have all the same problems discussed in part one e.g. Pausanias’ “Description of Greece” which praises Hadrian’s charity towards Athens appears to trace back to a single manuscript recovered by a Florentine humanist in 1418, which was of course immediately ‘lost’ despite the vast effort and expense that went into recovering it; although “Description of Greece” has enormous detail and citational value on many non-Hadrian topics it was virtually ignored by other authors throughout history (implying it may not have actually existed before 1418 at all).
Beyond a book riddled with “pure fiction” there is only one other piece of evidence that the wall was built by Hadrian: two small fragments of sandstone found in a church containing an inscription. But the inscription is so damaged that the reconstructed text is almost entirely guesswork.
From these two shards an entire text is summoned into existence through the magic of scholarship — the bold letters below are what’s legible, the rest is all invented by modern historians:
Divorum omnium filius
imp(erator) Caesar Traianus Hadrianus
Augustus imposita necessitate imperii
intra fines conservati divinopraecepto
… co(n)s(ul) III …
diffusis barbaris et
Britannia addidit limitem inter
utrumque Oceani litus per m(ilia) p(assuum) LXXX
exercitus p̣ṛovinciae opus valli fecit
sub cura A(uli) Platori Nepotis leg(ati) Aug(usti) pr(o) pr(aetore)
Most of the text and indeed most of Hadrian’s name is simply guessed, interpolated to fit the narrative from a story book. Who really built this wall? It’d be most accurate to say that we can’t say. But “Mysterious Wall” would be a rather boring moniker, and it’s a tourist attraction. So the spider-thin thread by which this narrative hangs is simply erased: neither schools nor Wikipedia explain just how unreliable the historical record really is.
The problems of scientific dating
Why do we so often find ourselves relying on dodgy books and fragmentary inscriptions? Hasn’t history been made scientific by rigorous dating methods, like radiocarbon dating or counting tree rings?
Scientific dating was supposed to revolutionise the study of history, perhaps even to partly automate it. No longer would scholars need to spend years piecing together complex theories about the dating of a dig site: just grab some pieces of biological material like wood or charcoal, send it off to a lab in a bag and presto. Back comes a precise date!
Yet some people still believe the timeline is wrong, so there must be caveats to these techniques. The historian Peter Jones discusses some in an FAQ accompanying a book he wrote on corrupted ancient chronologies:
- Radiocarbon dating often isn’t actually used when studying dig sites of ‘known’ age, because the consensus can’t be that wrong so why bother checking?
- When it is used and the returned dates conflict with scholarly consensus they’re just quietly shelved (this publication bias is a problem in all areas of science).
- It only works on biological samples like wood or charcoal, but there’s no reason to believe the ages of these things has to match the age of the site in question and usually they aren’t available at all, e.g. this is of no use in dating the wall.
- Scientific dating techniques don’t actually give dates in normal years. They give dates in forms like “radiocarbon years” or tree ring counts. To be usable they have to be translated to standard AD/BC dating, and it’s common for this mapping process to be calibrated based on other dating techniques —like things learned from old books. Sometimes this process is even circular. Thus a mistake in one can propagate to them all, whilst looking like they’re all reinforcing each other.
Jones’ FAQ says “the statistical variables involved [in calibration] are often poorly understood”. Cornell University found as recently as 2018 that these calibration curves are still wrong, even after decades of use of the technique:
“Our work,” he added, “should prompt a round of revisions and rethinking for the timeline of the archaeology and early history of the southern Levant through the early Biblical period.”
The amount of calibration is extreme for dates within the last few thousand years:
BP here means years before 1950 (“before present”) so a date of 1000 BP would be the year 950 AD. As can be seen the curve gets particularly wild as it moves further back into the post-Roman dark age.
Radiocarbon dates can easily be thousands of years too old. Decades after the introduction of the technique the age of freshly killed Antarctic seals was being “measured” as 1,300 years.
Anatoly Fomenko alleges further problems:
- Radiocarbon dating labs have a history of shoddy, un-scientific procedures that double down on circular propagation of mistakes. For example they may ask the submitter to provide an estimated date range, to help them check they got it ‘correct’.
- Labs often give very different dates for the same item. This does not appear to undermine archaeologist’s confidence in the technique.
- Because radiocarbon dating destroys the specimen calibration curves were calculated with a statistically meaningless number of samples.
Obviously such a technique can’t be relied upon to correct errors in the status quo understanding of history, only re-enforce it.
Dendrochronology (tree ring counting) has similar issues. Tree ring chronologies must be aligned with each other to stretch the ability to date further into the past, but the way this is done isn’t reliable whilst looking a lot as if it is:
An interesting problem with this sequence was revealed when Kuniholm tried to date wood from a city gateway at Tille Hoyuk. What happened was that the t-value wiggle-match produced by computer analysis came up with not one, but three matches of 1258, 1140, and 981 B.C. — each with a greater than 99.9% certainty. Also, dendrochronologists working in Hohenheim, Germany, were proven wrong three times in the mid 1990s, each time after very strong assertions of reliability.
Sometimes researchers prefer to believe things that defy common sense rather than accept they don’t know the truth. Bad datings are given ‘explanations’ rather than discarded.
In one notorious case a radiocarbon lab returned a date of 1000 BC for the bones of an Egyptian mummy but 300 AD for the cloth the mummy was wrapped in, a difference of over a millennium. Rather than reach the obvious conclusion that the very complex dating process might be buggy, they decided the mummy had been exhumed 1300 years later, rewrapped and then reburied for no reason. The mummy was selected for analysis specifically because it was so boring it didn’t even have a name, just a number, so it wouldn’t ever be put on display. The consequent implausibility of the reburial explanation was of no concern to the scientists who had invested so much time and reputational capital into their work.
Is history all wrong?
Dig into the origins of European history and an absolute chaos of competing stories emerge. Even when the subject presents itself as using the scientific method major problems get swept under the carpet. It’s not surprising that in this environment there are people who doubt the prevailing consensus, and even propose alternative histories.
These new histories may bear no resemblance to the ones we’re used to. For instance, one of the most interesting eliminates ancient Rome and Greece entirely: there was no fall of Rome or subsequent dark age. Everything we think of as “ancient” was in fact medieval in origin, and what happened before medieval times is unknown.
Does any of this matter?
It’s tempting to say no, not really. After all, who cares if the dark ages never happened, if ancient Rome didn’t really exist. They’re still all dead.
But alternative histories aren’t only good for entertainment. They remind of us how weak our grasp on historical events really is; useful to remember when faced with scholars telling us with absolute confidence what we can learn from history. They show us how quickly facts contradicting the narrative can be corrupted or covered up. When it’s about events that occurred thousands of years ago it’s easy to be detached, but studying these things gives us practice in spotting similar problems when they’re closer to our own time.
I’ll leave you with an example. Here’s an obscure quote by a man who lived in the 20th century.
‘We are socialists, we are enemies of today’s capitalistic economic system for the exploitation of the economically weak, with its unfair salaries, with its unseemly evaluation of a human being according to wealth and property instead of responsibility and performance, and we are all determined to destroy this system under all conditions.‘
This man led a movement that demanded, amongst other things, “all unearned income, and all income that does not arise from work, be abolished”, “the nationalization of all trusts” and “profit-sharing in large industries”. Although these are all classically hard left positions, our society believes by absolute consensus that he was on the far right. It impacts policymaking to this day.
Without looking it up, do you know who it was?