It is quite amazing, given the continuity of the speech signal, that we are able to understand what words are put together to form an utterance. This ability is even ore surprising because no two speakers ever say the 'same thing' identically. The speech signal produced when one speaker says cat will not be identical to the signal produced when another says cat or even when the first speaker repeats the word. George Bernard Shaw pointed to the impossibility of constructing any set of symbols that will specify all the minute differences between sounds, in his statement :
By infinitesimal movements of the tongue countless different vowels can be produced, all of them in use among speakers of English who utter the same vowels no oftener than they make the same fingerprints.
Yet speakers understand each other, because they know the same language. Our knowledge of a language determines when we judge physically different sounds to be the same; we know which aspects or properties of the signal are important and which are not. For example, if someone coughs in the middle of saying 'How (cough) are you ? a listener will interpret this signal simply as 'How are you ?' Men's voices are usually lower in overall pitch than women's; some speakers speak more slowly than others; some people speak with a 'nasal twang'. Such pitch or tempo differences or personal styles of speaking are not linguistically significant.
Our linguistic knowledge, our mental grammar, makes it possible to ignore non-linguistic differences in speech. Furthermore, we are capable of making many sounds that we know intuitively are not speech sounds in our language. Many English speakers can make a clicking sound, which writers sometimes represent as tsk tsk tsk, but these sounds are not part of the English sound system. They never occur as part of the words of the sentences we produce. It is, in fact, difficult for many English speakers to combine this clicking sound with other sounds; yet clicks are speech sounds in Xhosa, Zulu,Sotho, and Khoikhoi languages spoken in southern Africa just like k or t in English. Speakers of those languages have no difficulty producing them as part of words. Xhosa, the language name, begins with one of these clicks. Thus tsk is a speech sound in Xhosa but not in English; th is a speech sound in English but not in French. The sound produced with a closed mouth when we are trying to clear a tickle in our throats is not a speech sound in any language, nor is the sound produced when we sneeze.
The science of phonetics attempts to describe all the sounds used in human language sound that constitute an important subset of the totality of sounds that humans are capable of producing.
The process by which we use our linguistic knowledge to produce a meaningful utterance is complicated. It can be viewed as a chain of events starting with an idea or message in the brain or mind of the speaker and ending with a similar message in the brain of the hearer. The message is put into a form that is dictated by the language we are speaking. It must then be transmitted by nerve signals to the organs of speech articulation, which produce different physical sounds.
Speech sounds can be described at any stage in this chain of events. The study of the physical properties of the sounds themselves is called acoustic phonetics, and the study of the way listeners perceive these sounds is called auditory phonetics. Articulatory phonetics the study of how the vocal tract produces the sounds of language.
