Because most philosophies that frown on reproduction don't survive.

Saturday, October 08, 2011

On Reading History

Sometimes the posts I want to write most take the longest to get around to writing -- in part because writing is often a mode of problem solving, and the most interesting problems are often difficult to solve.

Case in point, a while back there was some discussion here about how to find good history books when I was ranting about Lied My Teacher Told Me and my issues with its take on US history, and a reader asked:
And so, where does that leave the rest of us who have only learned about history what we were taught in school...which it seems clearly can't be relied on.

If we don't already know what happened, how are we to know who (what texts?) to trust?

It's this kind of thing that makes me (and others) so frustrated as to maybe just not care. Find a book so you can learn more only to learn that you're (again) being taught the wrong things.
As I said at the time, this seemed like it deserved not just a comment in response, but a more lengthy consideration in a post. It's been rather longer in brewing than the result is likely to justify, but here it goes...

There is not, unfortunately, a magic formula or vitamin pill that can be taken in order to detect whether you should trust a given author or book. However, I think there are some general principles that will stand people in good stead.

Know How Much To Expect From Different Types of Books
History is a seriously huge topic. Considering the number of people, nations, tribes, etc. that have existed since people began recording their history in writing, and the amount of time and space covered by most popular history books, it's not surprising that the more general the history, the more cursory the presentation of events and trends. There's an idea out there that there is a pattern to history, and if only one can get a good account of the pattern and some of the major events, after that one will simply be filling in details. I would tend to say that this is mostly wrong. To the extent that there is a pattern to history, that pattern is simply the effect of a number of details which, together, seem to form a pattern. Patterns themselves do not drive history. As such, I think it's wise to expect that the more general the history book, the more the author will attempt to fill in gaps with narrative. And that narrative will often be, to the extent it is a simplification (and subject to the author's interpretations and biases) the less reliable part of the package. When you read a survey text that cover a large stretch of land or time (as "The History of the Americas" or "Medieval History" or "The Modern World") what you can probably rely on quite well is the timeline elements of the book: Who lived when, what they did, etc. What you should regard more tentatively is judgements which the author provides. So, for instance, if you're reading medieval history, you can believe the author as to when the First Crusade took place, who went on it, when various battles were fought, etc. Things you may not be able to rely on as well are general statements about tone or cause which the author makes: "The Crusaders were much less civilized than the Muslims they were fighting." "While they had a religious pretext, most Crusaders did not go on crusade for religious motives." "The Muslims were more less violent than their European adversaries." Authors may or may not be right in these kind of judgements, and as you read more about a topic, you'll come to your own judgement on these questions. But it's interpretive statements like these that will often be most affected by omissions or biases on the part of the author.

Usually, the more general the history (the longer the period or wider the area it covers) the more you're going to be getting the author's narrative interpretation of events, and the less detail you'll be getting about actual events. These kind of books are still quite useful. If written by a good author (and you'll eventually develop an instinct for detecting these) you may get a useful interpretive framework for looking at a period. But even from a poor author you'll get a good idea of what happened when and who the major players were. It will thus give you an idea where to read further if the topic seems worth learning more about.

Treat general histories covering big topics as timelines and frames -- read books focusing on a smaller stage to understand people and events in detail. As you get in to reading multiple books on a given topic, you'll see what authors are living issues out or being one sided in their presentation. The more background you have, the more you'll have the ability to glean useful bits even from biased books which at least present new information about one side of a topic.

Primary Sources Are Your Friend
If generalized narratives are your enemy, primary sources are your friend. A primary source is, quite simply, what someone wrote about a place or event at the time. Now, of course, people at the time are as fallible as anyone, and they can lie or have strong points of view, and an author can be selective in picking his primary sources. However, at a minimum, primary sources give you an idea of what some people at the time thought. If you can then seek out people on different sides of a given conflict or issue, and read primary sources from both sides, you now have a moderately balanced view of the situation.

Watch Out for False Narratives
So having warned of generalizations, what are some of the false generalizations or narratives to watch out for? There are some common ones:

The Myth of Progress
It's undeniable that there have been certain kinds of technical progress over time. However, one school of thought holds that people themselves and society as a whole have become significantly better and more noble over time. This is, to my mind, always a warning sign. If an author seems to be telling you that people in the past were all dumber or more wicked than people today, then he's probably not giving you a very balanced view of the topic he's covering. And, of course, the injustices of the past (to the extent that they are past) are often far more obvious and egregious to us than the injustices of the present -- since everyone knows that even if those really are bad, they can't be helped.

The Golden Age Myth
Somewhat less common at the moment (though in various quarters you'll see it, say, talking about the American Founders, Classical Greece, the Greatest Generation or the '60s), is the opposite narrative, that of the lost Golden Age from which all subsequent time has been a devolution. Just as the author who tells you that the present is wonderful and the past was terrible is probably mischaracterizing both the present and the past, so in the inverse is the case with the author who is convinced that some specific period in the past represents the Golden Age.

The Enlightenment Myth
Somewhat related to both of these is the Enlightenment Myth (thanks commenter Joseph M for the email suggesting this one) which combines the two. In this one, you have twin Golden Ages of reason in the Classical era and in the Enlightenment (both of these being rather loosely defined) with an abyss of superstition ruled over by the medieval Church in between. This usually combines a blindness to many aspects of Classical culture and the early modern "Age of Reason", and also a dismissal of the many achievements and virtues of Medieval culture. (Depending on the book and when it was written, you'll find this myth in both secularist and Protestant flavors -- the common thread being a desire to see a thousand year abyss of superstition, popery and bad hygiene between either Classical Greece and Rome and the French Revolution or the Apostles and Martin Luther.)

The Everyone Fits A Pattern Myth
This is a bit like the world building mistake that science fiction writer Orson Scott Card called "it was raining on Mongo that morning": Just because one has applied some sort of "spirit of the age" to a period (usually itself a suspect construction) does not mean that everyone then thought a particular way. Some like to call the Middle Ages the "Age of Faith" (this is, I think, a simplification for other reasons) and the next step is to then assume that everyone in the Middle Ages was incredibly devout. Then we have the Renaissance, when suddenly everyone loved the ancients and wanted to be an artist. Then there was the Age Of Exploration -- when suddenly everyone was curious about the world. And the Age Of Reason, when suddenly everyone was living by pure reason. You get the idea...

Counter Narratives Are Almost Always Heavily Biased
This, I think, is where Lies My Teacher Told Me especially falls down -- as do other polemical approaches to history such as The People's History of the United States. When an author sets out specifically to write a counter to what he perceives as an established and wrong narrative -- and does so by writing "the other side" rather than writing a fully balanced narrative of his own, the resulting book will likely only be useful to read as an afterthought to the narrative it's critiquing. A counter narrative is necessarily only half a narrative. (This is not the same as a "revisionist" approach, in which someone attempts to lay out a comprehensive narrative of his own which is different from a prevailing one. This can, at times, be very successful. For instance, some of the "revisionist school" of Israeli historians have done the most balanced, overall accounts of the Zionist movement and the early history of Israel. Some of the recent revisionist approaches to the history of the Great War have also been very good.)

Beware the Prosecutor Historian
A history professor once told me that the best history is written when a historian can get the prosecutor sitting on his should to be quiet. It's when historians are out to explain just who is at fault in some situation that they often end up letting contrary information slip -- information which is often more interesting than the "they were bad" kind of judgements which often take their place. One example that springs quickly to mind is a highly opinionated history book which was in some other ways enjoyable to me. In Modern Times Paul Johnson's explanation of Communist Russia is that the communists were "gangsters". Indeed, he essentially dismisses any ideological content to their activities. The problem is, this is actually far less illuminating about the very evils of the regime than looking at how their ideology and their utter lack of moral scruple interacted. Simply saying that Stalin was a gangster does little to explain the appeal that communism had in many parts of the world in the 20th century. It also left me wondering how much Johnson was simplifying other matters I knew less about. A good book seeks to lay out as much as possible about the people and groups involved, what they did and what their apparent motivations were, and leaves the reader to draw his conclusions with as little generalizing and editorializing as possible.


Jennifer Fitz said...

I like biographies, because the details of a person's life give evidence that can help fact-check historical narratives.

RL said...

Good stuff, Darwin. May I be so bold to add a couple more things that I think are important.

1. This expands a bit on one of your points, but is I think helps directly when choosing or reading a history. The book should have a well sourced bibliography. This not only gives you the option to dig deeper or to add a text to your reading list but indicates what sort of diligence went into the work. It seems like it may also be a good check for the author to avoid taking some unwarranted license.

2. I love it when a historian calls out matters where various historians disagree and then explains his position. It shows a degree of character and honesty, IMO.

Anonymous said...

A good post, but I have one quibble: progress isn't a myth. People really have become less violent over time. In America prior to Columbus it appears that 1 out of 6 people died violently, based on the number of skeletons that are found with arrowheads embedded in the rib cage or with huge cracks in the skulls. This astronomical murder rate appears to have been typical of hunter-gatherer societies around the world. Something about the rise of agriculture and urbanization has had a civilizing influence on people, for which we can all be thankful.


Brandon said...

People really have become less violent over time.

As John Gray keeps pointing out whenever Steven Pinker raises the reduction in violence point, there's good reason to think that peace in agricultural societies came in part by fomenting war among more nomadic societies (by playing them against each other one reduces their worrisome tendency to move against oneself; likewise, wars among the barbarians gives one something to do with ambitious potential rivals); and that peace in industrial societies has come in part by fomenting war among more agricultural societies (e.g., the proxy wars of the Cold War); and that the homicide rate in such local war zones sometimes exceeds 1 in 6 by an enormous amount; and that periods of peace are often secured by acts of intense violence beyond anything ancient tribes could have imagined -- e.g., one third of the entire thriving city of Hiroshima dead in a day, with twice that many dead of complications by the end of the year. In other words, much of the evidence that violence has declined could equally be read as evidence that it has not declined but merely been sequestered and quarantined in both place and time. And it is important in that, if the latter is true, what we really mean by 'violence has declined' is that violence we take to matter, namely, the kind that threatens ourselves, has declined, regardless of what violence threatens Those Other People.

Regardless of whether Gray's full argument is tenable, it makes the important point that artificial zones of lower violence don't actually tell us anything about whether people are really less violent until we have pinpointed their exact causes.

Anonymous said...

None of Gray's arguments applies to pre-Columbian America. And the fact that murder rates were similarly high in other hunter-gatherer societies suggests that his arguments are not tenable.


Darwin said...

From what I have read, I think that Pinker's numbers broadly bear out, despite Gray's critique.

That said, I don't think what Pinker is highlighting successfully shows that "people themselves and society as a whole have become significantly better and more noble over time." In a more affluent society, people have less of a tendency to resort to violence to solve their problems -- they have too much to lose. That doesn't, however, mean that people now are better than those in hunter gatherer societies. Their vices just tend to come out in other ways that resorting to casual violence, blood feuds, etc.

Take away that "something to live for" which modern material society suggests to people and things can become just as barbaric now as in the past.

Brandon said...

None of Gray's arguments applies to pre-Columbian America.

Gray's arguments are about what can reasonably be regarded as proven from the kind of data Pinker cites, not the data itself; his point being that such data points are useless if one has not made sure that standards of comparison are stable across populations and appropriate to the conclusions drawn. For instance, precisely one of the problems is talking about violent death rates as measures of violence as such, because such measures are susceptible of massive skew depending on size of population, and, indeed, on how, precisely, the population is defined. In a group of three, if one person is murdered the murder rate is 1 in 3; in a group of ten, if two people are murdered, the murder rate is 1 in 5; but in what sense is there less violence in the latter scenario? Killing 1 in 6 people out of 300 requires killing 50 people; killing 1 in 6 people out of 30000 requires killing 5000 people; killing 1 in 6 people out of 3 million requires killing 500000; killing 1 in 6 people out of 3 billion requires killing 500 million. As the population expands, the resources required to keep violent deaths at a constant rate massively expand. Because of this, talk of violent death rates is useless unless one has established that declining violent death rate is not due merely to the overall population outbreeding its resources for effective violence (particularly when combined with the sequestering noted before). There are six billion people in the world; 1 in 6 requires killing 1 billion people. Even world wars and the most genocidal civil wars can't get anywhere near that threshold. Further, consider a hypothetical situation in which there are no violent deaths at all: two nations pointing nuclear missiles at each other and constantly drilling war games, without any actual action against each other, only because of fear that the other nation will retaliate. It's a less violent situation, but does it really prove that the people involved have become less violent? Violence should not be conflated with violence that manages to be effective in terms of killing people; and it is not at all clear that even effective violence is measured best by a rate rather than absolute numbers or some other metric. And even more importantly, even if all that is overlooked, it is a serious fallacy to treat evidence that political situations have become less violent as a proof that human beings have become less violent.

Pinker may well be right in the end, but it needs to be recognized that this sort of reasoning is very much in line with Pinker's penchant for making very ambitious arguments for rather sweeping conclusions. The prudent person will be cautious -- particularly, as Gray occasionally notes, given that we are dealing with a conclusion that lends itself so easily to patting ourselves on the back.

Brandon said...

To put the problem in yet another way: In six years World War II wiped as many people from the face of the earth, and possibly more, than even existed in the pre-Columbian Americas, North and South combined, at the height of its population, and it did so in great measure from the advanced nations that Pinker lauds. So how again does the pre-Columbian violent death rate really show that we're getting less violent? Certainly not on its own; we'd have to look at underlying causal factors for what is going on.

I should say that Pinker's argument is actually pretty sophisticated, and that since he first started making it he has generally been very good at explicitly emphasizing the limits of both his data and, albeit not always so explicit, what his data show. He fully does recognize that he's making a very complicated argument for a conclusion many people are going to be skeptical of; my complaint is not so much with his argument, or similar arguments, as with attempts to short circuit the proper sort of evaluation such complicated arguments require before they can be trusted.

Darwin said...

I read Pinker's piece in the WSJ a couple weeks ago on this topic. Are there any more detailed magazine or newspaper pieces you'd recommend -- sort of turning to his whole book which is just out?

Brandon said...

He's been talking about it for several years now, actually, so I'm sure there are lots of things. (Pinker, despite his rather audacious arguments, tends to have excellent books, and I think it's precisely because he tests out his arguments in public for some time before booking them.) I saw a video lecture of him a while back in which he discussed it, and am fairly sure I've read some summaries as well as some bits and pieces of more recent things related to the book.

I can't remember if the TED video is the one I saw (four years ago!), but he has an essay at The EDge, and you might be able to find the relevant video at the TED site given at the link:

JMB said...

Good points. My brother just wrote a book about one of the founding fathers - Oliver Ellsworth. Check it out; it was reviewed in the WSJ a few weeks ago. (M.C.Toth)