daily in many newspapers across the country. In these puzzles, the player begins by analyzing visually-based linguistic clues, and then, by deduction, begins guessing which letter replaces the ciphertext letter in a trial-and-error manner. Examples of clues include the commonness of three-letter words such as the and and. Also, words with an apostrophe offer strong clues: the letter following the apostrophe could be an s, t, or m.

Substitution ciphers can be broken down into two subclasses: monoalphabetic and polyalphabetic. The above-mentioned cryptoquote puzzle is an example of a pure, monoalphabetic scheme. Here, each occurrence of the letter in the plaintext form is always replaced by the same letter in the ciphertext, in a one-to-one relationship [ 4]. There are certain inherent weaknesses in monoalphabetic schemes. One of these is that recurring letters make deciphering the encryption easier. For example, the word LADDER, with the two adjacent Ds in the middle, would yield ciphertext with two like, adjacent letters in the middle. This reduces the mathematical complexity of deciphering the message [ 4]. Substitution ciphers require some sort of mapping key in their implementation. Such mapping keys may be rendered in different forms that will be examined shortly.

Monoalphabetic ciphers are sometimes rendered in a “shift” fashion. That is, the message is encrypted by shifting backward or forward within the alphabet for the cipher-match by a given quantitative difference in alphabetical position. For example, let us say that the shift key is + 3. Then in this system, A = D, B = E, etc. To decipher, you simply move ahead (in a positive shift) the number of the shift along the alphabet. One of the more well-known renditions of this system from history is the Caesar-Shift [ 10]. Obvious weaknesses here include the ease of breaking the code, the need to have knowledge of the key at both ends, and transportation of the shift key. Another way to render a monoalphabetical cipher is to utilize a random set of charac-ter/symbol matches. Again, both sender and receiver need tables of matches, and transporting the map could present security problems for such a system. However, a random match cipher is stronger than a shift-cipher since a simple shift cipher can be broken with relative ease by repetitive effort until an intelligible result emerges. These types of ciphers can be thought of as code encryption at the character level.

A more sophisticated and robust means of cipher-based encryption is found in the polyalphabetic scheme. Here, letters or symbols are replaced by different symbols in the ciphertext by using more than one replacement alphabet. For example, if the name ABRAHAM were to be encrypted with a polyalphabetic system, the result could be something like BFGHURF. Notice that each occurrence of the letter A is replaced by a different letter. Some of the earliest known polyalphabetic schemes utilized hand-held, rotating disks and printed tables, including the famous Vigenère’s table, developed by the 16th century French cryptographer Blaise de Vigenère [ 10]. Figure 1 shows one of the more popular versions of Alberti’s disk, a rotating cipher disk from history [ 10].

The cipher disk consists of two disks with a common hub. There is an inner disk that rotates and a fixed outer disk. Characters on the two disks will match when aligned. Both sending and receiving parties utilize the same disk, and the setting of the initial alignment is agreed upon before communicating. This starting base is considered the key. A rotating cipher disk may be employed in either a mono- or polyalphabetic encryption scheme. In a monoalphabetic implementation, the disk simply maps one letter to another. A polyalphabetic scheme

Figure 1: Example of a cipher disk, Alberti’s disk.

 

may also be employed with the cipher disk. Instead of using the same alignment throughout the entire message, the polyalphabetic method calls for shifts in disk alignment at predetermined intervals. In this way, more than one alphabet is used in the process. There are as many mapped matches to a character as there are shifts. These ciphers are much harder to break than are the monoalphabetic versions.

The other major class of cipher is the transposition cipher. In this method, symbols are not replaced by others, but instead are altered in position within a block or word of plaintext. In a simple transposition cipher, the plaintext FIGHT could be rendered in ciphertext as TIFHG. Transposition ciphers utilize a mapping key that serves as a pointer to the cipher symbol. The first step is to set the message into blocks for mapping, then perform the encryption or decryption process on them [ 4]. Let us look at a simple mapping scheme:

ABCD

maps to

CDBA

The encryption algorithm, utilizing this map, or key, would transpose the plaintext BACD as DCBA and in turn, the decryption system would map in reverse to expose the plaintext.

As mentioned in the discussion of codes, there are encryption methods that utilize both code and cipher methods to create a layered encryption that is stronger and more difficult to break. Let us consider a simple example of a code-and-cipher method using the missive, “move to the west.” First, we encode the plaintext according to codebook entries.

 

plainword(s)
move
.
to the south
to the west

codeword(s)
gads
.
aeilthun
aeilbret

 

The encrypted message is gads aeilbret. Notice also that this code accomplishes some compression in the process by reducing the number of characters to be transported. Many times a compression facility is built into coding schemes. Now this encoded message can be