The first (and only?) biometric word list is one developed by Zimmermann and Juola, with the assistance of Zhahai Stewart and Grady Ward. It was developed to permit the verbal authentication of Diffie-Hellman key exchange, and of cryptographic key fingerprints. The information conveyed using a biometric word list can also be conveyed by simply reading decimal or hexadecimal numbers. However, some numbers sound similar to each other ("five" and "nine"; "B" and "D"; etc.) and can lead to errors in "transmission", so the conversion to words is used. It is similar to the use of the military alphabet for the communication of letters.
The Zimmermann/Juola list is actually composed of two parallel lists, one of two-syllable words, the other of three-syllable words. To compose the verbal equivalent of a numeric value, the two lists are used in an alternating manner; the first byte is represented by the appropriate word in the two-syllable list; the next byte by the appropriate word in the three-syllable list; etc. This serves as an error detection mechanism, to address the tendency of people to transpose consecutive words, duplicate words, and omit words when reading aloud lists of random words.
For example, some words from the parallel lists are as follows:
Byte | Corresponding words | |
---|---|---|
(hex value) | Two-Syllable | Three-Syllable |
20 | bison | butterfat |
29 | breakup | certify |
6B | glitter | Hamilton |
FE | woodlark | yesteryear |
38 | classic | consulting |
0D | ancient | asteroid |
D2 | standard | sensation |
Thus, the randomly chosen sixteen-digit decimal number 9052743058066898, equivalent to the fourteen-digit hexadecimal number 20 29 6B FE 38 0D D2 (with spaces added for clarity of demonstration) is represented by the seven-word sequence, "bison certify glitter yesteryear classic asteroid standard", as shown by the boldfaced words in the table.