There's a similar surge over a similar time interval for 'scary', so maybe it's just correlated with other properties of the content they're sampling (not unlik

There's a similar surge over a similar time interval for 'scary', so maybe it's just correlated with other properties of the content they're sampling (not unlike https://xkcd.com/1138/ https://xkcd.com/1138/): https://books.google.com/ngrams/graph?content=scary&year_start=1800&year_end=2008&corpus=15&smoothing=3&share=&direct_url=t1%3B%2Cscary%3B%2Cc0 https://books.google.com/ngrams/graph?content=scary&year_sta...