Solving cryptograms
Solving cryptograms
by Ajax
When working on an easy crypt, look for common words such as the, and, in, and of. Try to identify words with repeated letters, such as ABCADB (probably people, though it could be proper or a few other, less common words as well), or EFGGEH (almost certainly little), or a long word ending in -IJKK (-ness). Some solvers keep lists of the common “pattern words“ they encounter; books of pattern and nonpattern word lists are also available for the dedicated solver.
Letter frequency is a useful clue in easy crypts but less so in harder ones, where the message has probably been deliberately designed to avoid the usual frequencies. The most frequent letters in English are e, t, a, o, i, n, s, h, r, d, l, and u in that order (some studies have given slightly different results). These frequencies are only tendencies, not laws of nature; even in easy crypts, don’t expect to find this exact frequency. For harder crypts, the best approach is often to determine which letters stand for consonants and which stand for vowels. Experienced crypt- solvers feel that once they have that information, the rest is relatively easy. Many systems for identifying consonants and vowels have been developed over the years, some of them involving considerable calculation. The method given here is not the most powerful, but it’s relatively easy to use and a good one for beginners.
Spotting vowels and consonants
- Make a record of how many times each letter in the crypt is used, how many times the letter starts words, and how many times it ends words.
- Assume that any letter used only once or twice in the crypt is a consonant. Put an identifying mark around letters used once and another mark around letters used twice.
- If the number of times a letter occurs at the start and end of words is half or more than half the total number of times the letter occurs, assume it is a consonant. Underline such letters.
- If a letter occurs between consonants identified in steps 2 and 3, assume it’s a vowel. Put an X under assumed vowels.
- Two letters that reverse with each other (for example, when the combinations ‑JY- and ‑YJ- both appear in the crypt) are usually one vowel and one consonant. Other factors such as relative frequency can often indicate which is which.
- Marking vowels and consonants as suggested leads to the spotting of other vowels and consonants. For example, if a letter occurs next to vowels and starts or ends some words, it’s likely to be a consonant. Judge based on all the occurrences of a letter, not just one or two. Keep trying combinations of vowels and consonants until one arrangement seems to fit throughout.
Identifying letters
Now try to identify some letters in the cryptogram, and from these, go on to whole words. This usually takes trial and error, but here are some tips:
- When two consonants start a word, the second is often h, l, or r. You can distinguish h from the other two because it very rarely appears after a vowel and very often before one; l and r appear freely both before and after vowels.
- A consonant that often follows a vowel but seldom precedes one is often n.
- A vowel frequently found in third-to-last position is often i, as in -ing, -ion, -ive, and other endings.
- Consonants that end several words may be d, s, or t.
- Three consonants together at the end of a word may be -ght or -tch. Four consonants may be -ghts.
- Look for words that may represent the prepositions above, after, amid(st), among (st), behind, beyond, from, over, upon, into, or with. It’s very difficult to construct a message that avoids all prepositions.
- Look for prefixes like ex-, over-, un-, or up-.
- Look for suffixes like -ed, -er, -man or -men, or -ful.
- A low-frequency letter at the end of words may be y.
- The lowest-frequency letter among the assumed vowels may be y, as in sylph, nymph, lymph, hymn, myth, lynx, or pachyderm.
When you make a guess at one word, try out those letters in other words of the crypt. Once you have two words right, the rest of the crypt usually comes easily. Remember, keep guessing! What one mind can devise, another mind can decode.
Cryptograms: Other solving approaches
by Sibyl, et al.
Cryptograms are often deliberately constructed to outwit a single solving method: it’s not hard to write a message, for instance, in which Q’s and Z’s have higher frequencies than E’s and T’s. But it’s impossible to disguise every telltale characteristic of the language and still be writing meaningful English.
Titles can suggest words that may be in the text. The title of the Hindu nastika thumps mridanga… crypt above was “Atheism rewarded.“ If you look up “atheist“ in a large crossword-puzzle dictionary, you’ll find nastika and other words; trying nastika in all the possible positions will quickly lead to a solution. Similarly, faced with a hard crypt titled “Old fanfare,“ you might start by checking out the names for old trumpets.
Prepositions are the hardest words to hide. As Ajax suggests, try from, with, into, and so on.
Some solvers look for pairs of words with many letters in common.
Somewhere the crypt must have a noun plus a verb; therefore somewhere there’s a likely -s, either at the end of a plural noun or at the end of a singular verb (faun grabs or fauns grab). Or there will be a past-tense -ed. Constructors may disguise plurals and past tenses (children, seraphim, fish; brought, spent, came), but these variations are limited, and the disguises are penetrable.
Crypts tend to be limited in syntax. The first three words of a crypt, for example, are frequently adjective, noun, and verb (as in Hindu nastika thumps…)
[These articles originally appeared in the PDF of the Guide to The Enigma]