Google labs:word frequency in books over the last 200 years

The example in the OP (fuck) was so common until the early 1800s because of the typographic convention to substitute an f for an s. In other words, the word "suck" was being written as "fuck", which is why the word appeared so often until the early 1800s.

If you change the bounds to include the 1700's, the prevalence of the term is more pronounced (if you pardon the pun).

Any background on the origin of this typesetting convention? I'd like to know the whys and whatfors...

The technical term for this is a "long s", which Wikipedia describes at great length: http://en.wikipedia.org/wiki/Long_s

thanks, that was incredibly fascinating and enlightening, especially the examples.

The long s survives in elongated form, and with an italic-style curled descender, as the integral symbol ∫ used in calculus; Gottfried Wilhelm von Leibniz based the character on the Latin word summa ("sum"), which he wrote ſumma. This use first appeared publicly in his paper De Geometria, published in Acta Eruditorum of June 1686,[2] but he had been using it in private manuscripts at least since 1675.[3]

You can see the changeover quite clearly by comparing the two against each other: http://ngrams.googlelabs.com/graph?content=fuck%2C+suck&...

If we assume all pre-1800ish mentions of 'fuck' are definitely meant to be 'suck', it still features much more prominently in the corpus beforehand than after.

Any ideas why that might be? E.g. certain types of text that were more common before that era, or other (less, er, 'suck'y) types of text that came after, 'diluting' the corpus?

Utterly awesome. http://ngrams.googlelabs.com/graph?content=My+name+is+Inigo+...

Potentially even more awesome is that they have the entire dataset available for download o_O

edit: case sensitivity is more fun than insensitivity: http://ngrams.googlelabs.com/graph?content=Star+Trek%2Cstar+... vs http://ngrams.googlelabs.com/graph?content=star+trek%2CStar+...

edit2: there are a whole bunch of geek-term bumps around and just after 1900. Anyone know why? E.g.: http://ngrams.googlelabs.com/graph?content=Star+Wars&yea...

I have no idea, but my guess is that they don't know the dates for some books and the system automatically classifies the publication date as "1900" or "1901." If you search the word "quark," you also get a bump at around 1900 even though the word wasn't coined until Joyce's Finnegans Wake in 1939.

I find it kind of interesting that a lot of words peak around the middle of the 19th century and have been in decline ever since. I'm guessing this has something to do with the increasing number of books published but it is still kind of hard for me to imagine that "the" is less commonly used now than one hundred years ago. The pattern holds true for a lot of common words...

Is this weird ? :)

http://ngrams.googlelabs.com/graph?content=google&year_s...

http://ngrams.googlelabs.com/graph?content=terrorist&yea...

They had smartphones in the 1900s? Could this be related to that woman supposedly seen speaking on a cell phone in the Charlie Chaplin video?

http://ngrams.googlelabs.com/graph?content=smartphone&ye...

(Actually, "internet" also has a similar spike. I suspect some books are mislabeled in their dates.)

Some dates are mislabled, but mostly OCR errors:

http://www.google.com/search?q=%22internet%22&tbs=bks:1,...

Given the birther-related news today, I was curious about another uncouth term. Sad results:

http://ngrams.googlelabs.com/graph?content=nigger&year_s...

I don't know in what context they could use Geek in 1800. http://ngrams.googlelabs.com/graph?content=geek&year_sta...

Or the 1700s: http://ngrams.googlelabs.com/graph?content=geek&year_sta...

Perhaps weirder, "Woot": http://ngrams.googlelabs.com/graph?content=Woot&year_sta...

Seems like some are incorrect when i look at the scanned pages and one of those says Geck instead of Geek http://books.google.com/books?id=gvYqAAAAYAAJ&pg=PA497&#...

Ah, good point. This is generated by OCR + reCAPTCHA, error is guaranteed to creep in.

some interesting results: (sad about liberty)

http://ngrams.googlelabs.com/graph?content=liberty&year_...

http://ngrams.googlelabs.com/graph?content=l33t&year_sta...

http://ngrams.googlelabs.com/graph?content=hacker&year_s...

while love is going down http://ngrams.googlelabs.com/graph?content=love&year_sta... sad..

I'm going to have an impact on google's internet bill this month.