Monday, February 27, 2012

Playing with Ngram: Noticing Philosophical Trends

The other day I was playing around with Google's Ngram, which uses the books in Google Books to track the change in the use or popularity of different words or terms over time (basically how often a word or phrase appears in published works). Just out of curiosity I put in 'Plato' and 'Aristotle' to see how popular they have been over the years. The resulting graph was surprisingly interesting (you may have to click on the graph to see it).
Here Plato is in red and Aristotle is in blue. There are a few interesting things here. Before about 1790 Aristotle is hardly mentioned at all until his precipitous climb to popularity. I immediately wondered if this was an effect of the books they were using or if writers really didn't talk about Aristotle before the 1790's. Perhaps they referred to him by a different name, I know Thomas Aquinas simply called him "The Philosopher" (any insights from those readers who have more than a BS degree in Philosophy?). Maybe the English just didn't talk about him. But after about 1805 the two are almost inseparable, until about 1865 where Plato just seems to go off and have a party by himself, and Aristotle just tags along. They come back together around the end of WWI (this sounds like a bad soap opera), but then in 1960 Aristotle has his thing and Plato has never caught back up.

I also noticed that there was a striking drop in the popularity of both philosophers starting in about 1965 (with a strong revival 20 years later). I wondered if this effect could be seen in other philosophers. Behold the result:
Here I have included Aquinas, Aristotle, Plato, Nietzsche, Hume, Kant and Wittgenstein. David Hume died in 1776 (and we became a free country, coincidence? I think not! Definitely some causation there...sorry philosophy joke), so he was talked about even before he died. Kant died in 1804, but interest in in his stuff only slowly picked up until his first big peak in the 1880's. Nietzsche died in 1900, and had his first big showing during WWI. Wittgenstein died in 1951, which is when he started to get talked about (Lesson for today: If you want people to talk about you just have to die.).

I noticed that the decline of the 1960's was present in all philosophers except Nietzsche who had his second big bump (Wittgenstein did not drop but his upward progress was slowed). So now the big question, is this just an effect of how Google gets and processes the data, or did people really stop talking about all those philosophers in the 1960's? And what drove the revival in the late 80's and throughout the 90's, and now why have they all dropped off, including Nietzsche? Also, it seems that the period between 1860 and 1920 was more Platonic but since 1960 we have been more focused on Aristotle (Atlas Shrugged was published in 1957, which is almost exactly when Aristotle and Plato diverge. Coincidence? Probably not. And that is not a joke.). What effect has that had on us? A whole host of interesting questions.


  1. In 1960s we began our involvement is Vietnam.  Similarly, we went to war in Gulf beginning in 1990, and, began preparing for war of ideology which is still going on.
    Is it that when death rules, philosophy of and about existence is not needed? 
    Alternately, build of interest in existential blasphemy (philosophical justification for war) must reach to a peak before a nation embarks on death and destruction, as demonstrated by peaking in 1960 and 1990?
    QL42, you are on to something, possibly a gage to measure cultural phenomena of wars?

  2. There is a possibility there.  We see another dip in approximately 1910, roughly corresponding to WWI, but it doesn't seem to hold if you look at other major wars, like WWII (the 1930s and 1940s show a high on these charts).  The Civil War is too small to tell (it might be a dip, but I'd feel more comfortable saying it's within the noise of the data).  It personally seems to be part of a larger trend which started about 1820-ish and continued for most of the 19th century.  We see other major dips in popularity in the 1700s (mostly using Plato as the representative, due to lack of much of anything about the other philosophers), and I can't readily correlate these to any sort of conflict or social unrest. 

    I would personally posit an alternative hypothesis.  The 1960s saw a significant social rejection of long-held philosophies.  It seems rather intuitive that the generation that brought us the "social experiment" of the 1960s would be less interested in the ideas of Plato and Aristotle, as they sought their own, modern philosophies and philosophers.  If you subscribe to William Strauss and Neil Howe's generational theory of history, there would seem to be a stronger correlation between the dips and the rise of the "Nomad" or "Artist" archetypes.  (I haven't spent too much time trying to compare the two, and I might be confusing the architypes, but the correlations seem roughly right over the last century or so.)  Again, the comparison isn't perfect, and Strauss and Howe's ideas aren't universally accepted, but it does seem to reflect the changing social attitudes over the years. 

    In any case, it's a very interesting trend.  Anyways, QL42, as always, great post. 

  3. I think perhaps a more significant correlation here is that the Beatles became popular at about the time that traditional philosophy took a dive. The effect was so strong that it took until the 90's to recover. Philosophy then reached it peak again but was undermined by the release of the Beatles' Anthology in the mid 90's.

    Philosophy, just can't compete with pop music.

  4. We discussed in my philosophy class today that Aristotle's writings were lost for much of the Medieval Age, and he mostly resurfaced around the time of the Renaissance. I thought that was rather interesting.

  5. How the writings of Aristotle returned to Western Europe is a long and complicated history. After the fall of the Western Roman Empire all, or nearly all, of the writings of Aristotle were lost in Western Europe. The problem was that they were all written in Greek and had never really been translated into Latin. The Western Empire only spoke Latin, while the Eastern Empire spoke Greek. So even if they had writings of Aristotle in the West, no one could read them, so they were all "lost". In the Eastern Empire in Constantinople they still had the writings of Aristotle and were able to preserve them.

    The writings of Aristotle were also preserved by Persian scholars who created a body of commentary, but after the 5th century there were effectively no one in Western Europe that could read Greek, thus Aristotle was only preserved in Eastern Europe and in Muslim countries, where there were still Greek speaking people.

    In 711 the Moors invaded Hispania, and it is assumed that some time later, it is not known when, they brought the writings of Aristotle back to Western Europe. When the Reconquista began in the 11th century these manuscripts began to slowly circulate in Western Europe, but still there weren't a lot of people that could read Greek so it was limited to a very, very, very small group of people in Western Europe (keep in mind that during this time the writings of Aristotle were widely read and used in the Eastern Roman Empire, commonly called by us (incorrectly) the Byzantine Empire).

    Also in the 11th century the first of the Crusades began when the Eastern Roman Empire asked the Pope in Rome for assistance in fighting the Turks. This brought many Western Europeans to Constantinople, where they still spoke Greek, and allowed some of the Crusaders to learn Greek and pick up on Greek and Roman literature that had been largely lost in the illiterate West. As far as I know, when Aristotle was introduced back into the West is unclear, but his writings were extensively known and used by scholars in the first Western universities, to the point that when St. Thomas Aquinas wrote his Summa Theologica in the 13th century he referred to Aristotle simply as "The Philosopher" because next to him (Aristotle) there was none other (Plato got mentioned, but he was merely a warm up act to the main event, quite the opposite to how Plato is viewed in modern philosophy).

    So by the time the Renaissance came around Aristotle was well known and had been used in scholarly writing for several hundred years.

    Upon further investigation it would seem that the first translations of Aristotle into English came in the 16th century. But his complete works were not translated until the 18th century, with the complete works in the 19th century. So that may account for the sudden rise in the end of the 18th century.

    But as I mentioned in the post, this could be an effect of the data since Google's Ngram does not have many books before 1800.

  6. I like that!  Yet, Lennon's message in his song "Imagine" was not listened or understood.

    Honestly, I run when we start pounding chests and justify our pounding through esoterics.

  7. Would you say that N-Gram statistics are at fault as we go back in time?  Published material exploded post WWII and after the Internet.  Unless ofcourse Google adjusts for such stuff.  I subscribe that anarchy rules under the disguise of democracy, theocracy, kingdoms, dictatorial rules, and any other mname you may want it to call.

  8. "Would you say that N-Gram statistics are at fault as we go back in
    time?  Published material exploded post WWII and after the Internet. 
    Unless ofcourse Google adjusts for such stuff."

    Well, it's kind of hard to say without a close look at how Google's algorithms work.  Are we looking at essentially a percentage of published works (or even philosophy works) talking about Aristotle (for example) or a total number talking about Aristotle?  If we were going to be really honest here, all of those lines would come with error bars which would, likely, grow significantly as we go back in time. 

    However, if we mostly just compare time periods that are fairly close together (i.e. we don't try to compare the early 21st century with the early 17th century directly) so that the amounts of uncertainty don't change too much between compared points, and we look for general trends, I think the conclusions would still hold even without dealing with all the statistics. 


To add a link to text:
<a href="URL">Text</a>