Obfuscator encodes arbitrary binary data into the case of words in an output text stream. The first four letters of each word are used for encoding; remaining letters in each word, if any, are not significant in the encoding. Words in the word list which are shorter than four letters may appear in the output, but are not used in the encoding. The total output is a whitespace-delimited list of words, within which the case of letters varies without regard to grammatical standards or common usage norms.
An encoding program may select and sequence words for output without regard to grammatical or syntactical sense of those words as linguistic text. Encoding programs may create "word salads" from internal word lists, create words from random letter sequences, use base text documents or multi-word text fragments, or some other method resulting in whitespace-delimited words containing at least four characters capable of case variation to carry the obfuscated data.
The technique of disguising a message within seemingly random variations or garbage areas of another message is properly referred to as steganography. When an otherwise reasonable text document is used as the base text for encoding a message via case variation, this technique may be considered a form of steganography.