Hypercorrection is an interesting word. Is there a way for something to be more correct than right? Of course this is not what is meant by hypercorrection, but hypercorrection means the correction of mistakes and transferring the new rule and applying it to correct utterances. Therefore this will appear as a mistake but because of the new rule of the performer he or she will not recognize it. This way of ``correction'' is therefore called hypercorrection. Mentioning this in the context of language change might result in the question why this is necessary. The reason is obvious: Hypercorrection is a result of a misinterpreted or misapplied rule. Therefore the rule of the speaker seems to be changed in some way; he or she uses it regularly. As previously discussed (see p. 
) this person might be a new standard, a new reference for others. According to the previously discussed factor of prestige (see p. ) this hypercorrection has different results according to the group of people and their social status.
According to Labov [Lab66] an interesting phenomenon appears in the lower middle class: The middle class uses the upper-middle class and/or the upper class as their reference but as part of their own society certain pronunciations do not appear in casual speech or at least not as often as in their reference group. The more formal a sitation is the more will the middle class adopt the features of the higher class and will even overtake the rate of the upper class in the most formal style. This is the result of the higher prestige of the accent of the higher class. The higher class is not in a process of changing the same features and therefore does not ``correct'' their pronunciation. They are in a state of stable affiliation to their class while the lower middle class strives for social advancements and is therefore in a state of changing affiliation.
Up to now it still seems to be bare theory that language change happens by hypercorrection, but according to Labov it is an evident process. He refers to the constriction of /r/ which became part of the New York City accent after World war II [Lab66]
. Labov describes the process such that a child up to 2 or 3 years will be parent oriented in his or her accent, and later, up to the teenager years their peer group will have a much larger influence on them. Beginning at the age of 16 or 17 a teenager is able to understand the usage of prestigious accents and learns to use them. Someone from a higher social class is more likely to use the more prestigious accent because he/she has had contact with many more speakers of the more prestigious accents and by this feels secure in using this accent. Someone from a lower class will lack this security because of less contact with speakers of the more prestigious forms. If these lower class speakers use the more prestigious forms they will do it, due to their lack of security, more consciously and therefore more correctly than the -- natural -- speakers of the upper class accent. This might be concidered as hypercorrection then.