Codes, ciphers, and cryptography à la chinoise et à la japonaise

« previous post | next post »

This is a passage from chapter 3 of Dan Brown's Digital Fortress (1998)

Eventually one of them [VHM:  NSA cryptographers] explained what Becker had already surmised. The scrambled text was a code‑a “cipher text”‑groups of numbers and letters representing encrypted words. The cryptographers’ job was to study the code and extract from it the original message, or “cleartext.” The NSA had called Becker because they suspected the original message was written in Mandarin Chinese; he was to translate the symbols as the cryptographers decrypted them.

For two hours, Becker interpreted an endless stream of Mandarin symbols. But each time he gave them a translation, the cryptographers shook their heads in despair. Apparently the code was not making sense. Eager to help, Becker pointed out that all the characters they’d shown him had a common trait‑they were also part of the Kanji language. Instantly the bustle in the room fell silent. The man in charge, a lanky chain‑smoker named Morante, turned to Becker in disbelief.

“You mean these symbols have multiple meanings?”

Becker nodded. He explained that Kanji was a Japanese writing system based on modified Chinese characters. He’d been giving Mandarin translations because that’s what they’d asked for.

“Jesus Christ.” Morante coughed. “Let’s try the Kanji.”

Like magic, everything fell into place.

David Becker is:

…The youngest full professor at Georgetown University and a brilliant foreign‑language specialist, he was practically a celebrity in the world of academia. Born with an eidetic memory and a love of languages, he’d mastered six Asian dialects as well as Spanish, French, and Italian. His university lectures on etymology and linguistics were standing‑room only, and he invariably stayed late to answer a barrage of questions. He spoke with authority and enthusiasm, apparently oblivious to the adoring gazes of his star‑struck coeds.

No matter how smart David Becker is, the above passage left me reeling and roaring with laughter.  As is true of almost everyone on the face of the planet, Dan Brown doesn't seem to have a clue as to how the Chinese and Japanese writing systems work and how are they are related to each other.

Referring to Chinese characters as "symbols" right away gives one cause for grave reservation.  When Becker tells the NSA cryptographers that the Mandarin "symbols" had the "common trait" of also being "part of the Kanji language", the commotion in the room instantly falls silent at this amazing profundity.  When the leader of the team, "a lanky chain‑smoker named Morante", turns to Becker and asks in disbelief, "'You mean these symbols have multiple meanings?'”, this is what happened to me:  🤣

It only gets better / worse. Becker explains to the breathless team of NSA cryptographers that Kanji is "a Japanese writing system based on modified Chinese characters", and that he'd been giving the team Mandarin translations because that's what they asked for. Whereupon lanky, chain-smoking Morante coughs and says, "Let's try the Kanji". Becker tries the Kanji, then, like magic, everything falls into place.

Brown portrays the NSA cryptographers as not knowing the difference between language and script and being unaware that the distinction between Chinese and Japanese is not merely a matter of flipping a switch between a kanji system and a hanzi system.

If that were the case, woe is us.  Thankfully, it is not.


There is probably no subject on earth concerning which more misinformation is purveyed and more misunderstandings circulated than Chinese characters (漢字, Chinese hanzi, Japanese kanji, Korean hanja) or sinograms.
–Victor Mair
from the foreword to Ideogram, by J. Marshall Unger

See also:


Selected readings


Additional Language Log posts about Dan Brown's novels and related topics:

Also here.


  1. David Cameron Staples said,

    January 8, 2021 @ 8:54 pm

    Dammit, you're destroying my illusions! Dan Brown doesn't know enough about linguistics to aspire to even being wrong?

    Next you'll tell me that he's authoritatively not even wrong about medieval history, renaissance art, nuclear physics, or nanotechnology engineering, or whatever "symbology" is supposed to be!

  2. Bathrobe said,

    January 8, 2021 @ 9:15 pm

    Of course, if it were in Japanese it wouldn't just have kanji….

  3. Victor Mair said,

    January 8, 2021 @ 9:42 pm

    Language Log knew about this formulaic Dan Brown novel long ago:

    "A five-letter password for a man obsessed with Susan" (9/10/05)

  4. Jim Breen said,

    January 8, 2021 @ 10:57 pm

    I started reading. Digital Fortress about 20 years ago. As a former employee of a communications intelligence agency I thought it might be interesting. I quickly abandoned it; it was wildly and hilariously inaccurate at every turn. I never got as far as the "kanji language" passage.

  5. JB said,

    January 8, 2021 @ 11:18 pm

    Call off the search for Athanasius Kircher's reincarnation.

  6. Lukas said,

    January 9, 2021 @ 2:37 am

    "Dan Brown doesn't seem to have a clue as to how the Chinese and Japanese writing systems work"

    Dan Brown also doesn't have a clue as to how cryptography works, so two wrongs make a very wrong.

  7. Joyce Melton said,

    January 9, 2021 @ 3:09 am

    I don't think I got past page 8 of this book. Someone else in the agency I used to work for asked me about it and I said, "Dan Brown may actually have been to Bethesda at some time because he described the front of the building correctly."

  8. Philip Taylor said,

    January 9, 2021 @ 4:52 am

    ∃ at least one reader of Language Log who has neither read nor intends to read any book by Dan Brown. I state this from personal knowledge.

  9. F said,

    January 9, 2021 @ 4:59 am

    But Brown is (apparently) not interested in correctness: merely in writing something his chosen audience (people who don't know anything about anything, perhaps) will buy. I gather he is quite good at that.

  10. Brandon Seah said,

    January 9, 2021 @ 8:47 am

    “You mean these symbols have multiple meanings?”

    Becker nodded. He explained that Rechtschreibung was a German writing system based on modified Italian characters. He’d been giving German translations because that’s what they’d asked for.

    “Jesus Christ.” Morante coughed. “Let’s try the Italian.”

    Like magic, everything fell into place.

  11. Ken said,

    January 9, 2021 @ 9:12 am

    This does interest me in how Chinese and Japanese cryptography was done in the pre-digital age. I'm sure google will turn up any number of sources, but can anyone suggest a good one?

  12. Not a naive speaker said,

    January 9, 2021 @ 1:38 pm

    David. Kahn the Codebreakers is a good starting point.
    Everything you want to know about codebreaking up to the 1960s

  13. Peter Taylor said,

    January 9, 2021 @ 2:16 pm

    @Ken, the similarly named Codebreakers (ed. F. H. Hinsley and Alan Stripp, ISBN 0-19-820479-5) describes the work done by British cryptographers at Bletchley Park during WWII, including some work on Japanese codes and ciphers, and describes the most relevant ones. Summarising, the lower level codes were simple code books, using four or five kana to represent each word/digit/romaji letter, although to try to avoid it being too easy to tackle by frequency analysis there were many alternative sequences given to encode common symbols. Higher level ones used the linear addition of a key (multiple-time pad, subject to the same vulnerabilities as VERONA). Merchant shipping used the kind of simple transpositional ciphers that I played with in primary school.

  14. Tom Dawkes said,

    January 10, 2021 @ 1:35 pm

    It's also infuriating that Brown refers to "six Asian dialects". He might as well have written "six Asian dialects [namely, Chinese, Japanese, ………and three European dialects [namely Spanish, French, and Italian]".

  15. Victor Mair said,

    January 10, 2021 @ 2:15 pm

    @Tom Dawkes


  16. Vanya said,

    January 11, 2021 @ 3:44 am

    three European dialects [namely Spanish, French, and Italian]

    Conversely, a scholar who has mastered Lebanese, Egyptian and Gulf Coast, MSA and classical Arabic would probably just get credit for “mastering Arabic”.

  17. Victor Mair said,

    January 11, 2021 @ 8:16 am



  18. ohwilleke said,

    January 20, 2021 @ 5:27 pm

    I would agree with almost all of the criticism of Dan Brown's novel except the use of the word "symbol" which is a Dan Brown "mythos" conceit.

    In a concession to the fact that his average reader doesn't know what "semiotics" means and is too lazy to look it up, some of Dan Brown's heroes are "symboligists", a fake discipline that is as normal as economics or psychology in the world of his books that is a bit linguistics, a bit anthropology, a bit mythology, and a bit genuine semiotics (because heroic archaeologists has already been done).

    I suspect that the word "symbol" in that passage is calculated to reflect that "Dan Brown World" conceit somewhere down the line, although possibly in its embryonic form as Digital Fortress is one of his first novels and he didn't have it all worked out properly at that point.

RSS feed for comments on this post