fbpx
The Stack Archive

Mark Zuckerberg’s bad karma with spell-checkers

Mon 21 Sep 2015

Mark Zuckerberg

While running over this morning’s post about Facebook’s ‘Dislike’ button being leveraged by phishers, I routinely right-clicked the word ‘Zuckerberg’, since the British English dictionary in Firefox was underlining it in red. The word was spelled correctly – I just wanted to access the contextual menu in order to ‘whitelist’ the supposed misspelling and lower the signal-to-noise ratio of words that the British English dictionary might highlight in future.

I have to admit that the dictionary’s suggestion seemed a bit of a reach, based on very limited word-parity with the flagged ‘Zuckerberg’:

zuckerberg-cocksucker

Curious, I tried out some more names that begin with ‘Zucker’, but they don’t seem to draw the same level of invective:

zucker-variations

I tried switching to the U.S. English dictionary, to see what its opinion was of the founder of Facebook:

zuckerberg-rubbernecker

Zuckerberg has suffered from this lexical affliction since around the middle of 2011, as far as I can tell, almost immediately after he was taken to task in the media for the disclosure that he had called Facebook users ‘Dumb fucks’ as a callow youth. But it seems to be a coincidence.

Though initially reported as a suggested substitution in Google Chrome’s spell-checker, the suggestion ‘Zuckerberg/cocksucker’ appears for me in the latest version (40.0.3 at the time of writing) of Firefox with the British English Dictionary pack installed.

The word ‘cocksucker’ does not appear anywhere in the .jar file inside the .xpi, and neither does ‘Zuckerberg’. Apparently it is not common OS source code that leads to the suggestion but is more likely to be the implementation of pattern matching under the Soundex and Metaphone phonetic comparison algorithms, as used across spell-checking APIs, including Google’s own implementation, which in Chrome does not even flag ‘Zuckerberg’ by default, unless you forget the capital ‘Z’.

The algorithms in question codify consonants with numeric values, giving each individual word a ‘hash’-style signature which exposes the rudiments of its array of consonants and vowels. Sometimes the match can be unfortunate.

Furthermore, since cocksucker is not a word the British English (and also the Australian English) Dictionary extension has in its database on install, the end-user is apparently only going to see this libellous offering if they add the word to their custom dictionary at some point:

add-to-dictionary

So it seems that’s what you get for being enough of a grammar Nazi to run a spell-checker during a flame war.


PS – On pasting this article into the CMS, I got the red line again in Firefox. This man just cannot win:

zuckerberg-knickerbockers

Send us a correction about this article Send us a news tip