After watching a TED talk entitled “What We Learned From 5 Million Books”, I may have found a brand new favorite thing on the Internet.
If you don’t have 14 minutes to watch the TED Talk, here is the summary: Engineers took Google’s digitized library and starting mining it for data. Essentially they have created a way to search words and phrases for 5 million texts stretching back to 1500. They claim that there are more than 500 billion words in the data set. Granted, the data is only from books – there are plenty of other cultural sources like magazines, newspapers, websites, art, maps, etc. Until those things are digitized and made searchable, we have the NGram Viewer, free, courtesy of Google.
Here are some interesting searches:
Fig. 1: Tragic Heroes from Shakespeare
Shakespeare’s tragic heroes generally trend upward, and Hamlet is the clear winner after some challenges by Macbeth and Lear in the late 18th century. Romeo has a steady if modest appeal. Macbeth dips below Lear in the 1960s, although that period marks a bear market for tragic figures stretching to 2008, the last year for which there are data.
Fig. 2: Curiosity, wonder, awe
An interesting trend line for “curiosity”, which is almost nonexistent until the middle of the 18th century (the rise of science?), has its hey-day in the early 19th century, and has been on a slow decline ever since. “Wonder” remains relatively popular, while “awe” remains stable at the low end of popular consciousness.
Fig. 3
The political landscape shifts in the 1980s: George Bush (senior) is the Vice-President in 1981, previously serving as a congressman, governor and the Director of Central Intelligence, and Clinton, Gore and Perot begin their careers. Clinton takes off from there, overtaking Bush (senior plus junior, since I can’t get it to recognize HW vs. W) at around 2001. Perhaps this reflects a spate of publications that reflect on the Clinton era, and perhaps his name is mentioned frequently in connection with his wife, Hillary. The decline in George Bush is strange since he is the current president until 2008, the last year of the data set. Gore has a steady, sauntering upward slope, and Perot is faltering. What ever happened to that guy?
Fig. 4: Philosophers, American English
Freud clearly dominates after 1940. Plato does little better than Nietzsche in the end, and Aristotle slightly better.
Fig. 5: Philosophers, French
The same time period, but from the French corpus. Similar upswing in Freud, although it doesn’t happen until about 1950, and Aristotle gives him a run for his money the entire time (spelled “Aristote” in French). Plato keeps the rest at bay, but Sartre is edging closer these days. Nietzsche and Heidegger, neighbors to the east, are far more popular than in English. In general, these names account for about twice as many words by percentage than in English (Freud peaks at 0.004% in French, at 0.0022% in English.)
Fig. 6: Philosophers, German
A very different picture in German. Freudian dominance is erased. Aristotle (spelled “Aristoteles” in German) is the strongest consistent winner, and Plato fades after 1860. Freud rises, but nothing meteoric like in French or English. Nietzsche, Hitler’s favorite philosopher, actually overtakes Aristotle in the 1940s, and does it again in the 1990s. Heidegger peaks in the ’90s as well. Similar to French looking at the percentage of these names in the entire German corpus. The Germans like their philosophy.
Taking the three pictures together, it appears that there is general recession in philosophy after about 1990.
Related posts:
