In sociolinguistics, SPEAKING or the SPEAKING model, is a model socio-linguistic study (represented as a mnemonic) developed by Dell Hymes. It is a tool to assist the identification and labeling of components of linguistic interaction that was driven by his view that, in order to speak a language correctly, one needs not only to learn its vocabulary and grammar, but also the context in which words are used.

Speech is the vocalized form of human communication. It is based upon the syntactic combination of lexicals and names that are drawn from very large (usually about 1,000 different words) vocabularies. Each spoken word is created out of the phonetic combination of a limited set of vowel and consonant speech sound units. These vocabularies, the syntax which structures them, and their set of speech sound units differ, creating the existence of many thousands of different types of mutually unintelligible human languages. Most human speakers are able to communicate in two or more of them, hence being polyglots. The vocal abilities that enable humans to produce speech also provide humans with the ability to  sing.