Helsingin yliopisto
Historians deal with such vast sets of data that is very difficult to keep oneself from prematurely reaching conclusions, especially if the research interest was motivated in the first place by a pet hypothesis. Our assumptions guide us even before we have decided on a single source to study, and this process tends to start feeding itself the further we go on and the more pieces of evidence we find in favour of either our hypothesis, or the first plausible narrative that we latch ourselves onto.
Yudkowsky calls those who write down the bottom line first clever arguers as opposed to curious inquirers.¹ The clever arguer begins by writing the conclusion they wish to reach, and then begins to accumulate evidence and construct arguments in support of their preferred bottom line. This is also called rationalization, which has nothing to do with making the process of rationality. As far as cognitive biases go, it is one of my ‘favourites’ since it is so ubiquitous in everyday life. Yudkowsky would prefer to call rationalization “giant sucking cognitive black hole” to avoid confusion between it and rationality. Meanwhile, curious inquirers first gather all evidence before writing down their conclusions
As no one actually starts by writing down the bottom line physically and historians usually do spend quite a large amount of time going through sources both in favour and against their hypothesis, it is easy to flatter ourselves by considering ourselves curious inquirers. However, critically looking at our own thought and decision processes, it is easy to see this is not often so. And though it is bad, it is not the end of the world.
“Most legal processes work on the theory that every case has exactly two opposed sides and that it is easier to find two biased humans than one unbiased one. Between the prosecution and the defense, someone has a motive to present any given piece of evidence, so the court will see all the evidence; that is the theory.”²
This seems to be the way science is mostly done, especially in the humanities. Even though it would be better for people to be curious inquirers, it is good enough that we have enough biased arguers – as long as they are on opposing sides so that the reader may find all relevant pieces of evidence between them. The problem with this situation is that people are unlikely to seek out information that would disprove their own hypothesis, so instead of academic readers getting many sides of the argument, they usually just choose a camp and choose to read and cite articles from their cite accordingly, ignoring the other side’s contributions. And as much as that is partly on them, wouldn’t you want to be one of those curious inquirers who is not just getting cited by people who already believed your point before reading a word of your research, to prove their own point? Or do you want to be a tool?
In hard sciences, it is easier to counteract this effect of being drawn into being a clever arguer when hypotheses can be tested with novel experiments. It is easier for a scientist to change their mind or be driven by curiosity when there’s new evidence to be found entirely. Alas, with history not only can we never test our hypotheses, new sources are very rarely found that would carry enough significance to shake anyone’s convictions. In reality, all historical conclusions are probability estimates with no way of empirically testing whether the estimate is correct or not. The validity of historical arguments is in practice measured by how convincing a case the historian can make, and most people are not in a position to notice if key pieces of evidence are being ignored by a clever arguer. If you decide which sources to include based on whether it is favourable or unfavourable towards the narrative you are trying to construct, you have become a clever arguer. Yet it is so tantalizing to do this with historical narratives as it is exactly what one is expected to do with all creative forms of narrative entertainment. When writing compelling fictional narratives, you are supposed to subtract what doesn’t drive the plot forward, and you should be as consistent with the alignment of your themes and messages as possible to make the experience enjoyable and approachable for the reader/viewer of narrative stories.
So why not just succumb to the allure of becoming a clever arguer like most everyone else?
Why, because at least I didn’t really get into academics to become a tool, or the ‘second best solution’ in conducting academic debates. I’d rather my research stand on its own rational merits, even at the expense of the overarching narrative.
“You cannot obtain more truth for a fixed proposition by arguing it; you can make more people believe it, but you cannot make it more true.”³
I for one do not like to form hypotheses before I set out to study something – it is evident that I have a general idea of what I am going to find, but at least so far I have not been very attached to my pre-conceived notions and am just curious to find out what my sources reveal. Not every change is an improvement, but every improvement is necessarily a change. If I were to reach the same conclusions after I had spent N amount of time doing research as the one I had when I started, I would be incredibly disappointed.
¹ Yudkowsky, Eliezer. ”The Bottom Line” in Rationality: from AI to Zombies” Berkeley, MIRI (2015). 302–304.
² Yudkowsky, Eliezer. ”What Evidence Filtered Evidence” in Rationality: from AI to Zombies” Berkeley, MIRI (2015). 307.
³ Yudkowsky, Eliezer. ”Rationalization” in Rationality: from AI to Zombies” Berkeley, MIRI (2015). 310