In many varieties of Southern States English, words containing a coronal stop such as 
and sometimes 
followed by a high back vowel 
have a glide 
inserted between the stop and vowel. Therefore, words such as tune, duke and news are pronounced as 
, 
, and 
. The forms with 
are most commonly glided, with 
less common, and even more rarely. The forms in some varieties become the affricates 
(eg. 
).
The background of this feature is uncertain, however one theory is that these glides are a remnant of an earlier British English pronunciation of the vowel as 
.
In Phillips (1994), the author describes a study done on the decrease in the usage of post-coronal glides in Southern States English. A previous study, Phillips (1981), had collected data from 18-22 year olds in Georgia. The informants were to pronounce a list of words that included words with the combinations tu-, du- and nu-. The words in the study varied in frequency.
As the table shows, the most infrequent words, such as Tudor, duly, and nude have a high percentage of informants without a post-coronal glide (87%, 69%, and 82%, respectively). On the other hand, more frequent words, such as Tuesday, new, and due, have a significantly lower percentage of informants without the post-coronal glides (51%, 43%, and 55%, respectively). Phillips (1981) takes this data to be an indication of a sound change in progress, where the change from a pronunciation to a pronunciation is occurring in the least frequent words first.
Phillips (1994) expands this study, using data collected from Telfair County, Georgia. The informants in the study were from two age groups: 13-19 year olds and 66-88 year olds.
Table 2 shows that for the 66-88-year-olds, the percentages again show that the most infrequent words have a higher percentage of glideless pronunciations than the more frequent words.
Table 3 shows that for the 13-19-year-olds, the percentages once again show that the most infrequent words have a higher percentage of glideless pronunciations than the more frequent words. However, in contrast to the older speakers, the percentages of speakers with glideless pronunciations is higher overall, indicating that the older speakers are more conservative, preserving the glide in more pronunications.
Also, it is clear from the data, summarized in the table below, that the younger speakers do not have as drastic a difference in percentages between most frequent and least frequent words as the older speakers.
| Table 4
Comparison of the Two age groups |
||
| Frequency group | 66-88-year-olds | 13-19-year-olds |
| 0-01 | 59.4% | 69.8% |
| 01-10 | 39.3% | 73.3% |
| 11-100 | 19.0% | 54.1% |
| over 100 | 11.3% | 53.0% |
Phillips takes this discrepancy as evidence that this is a sound change in progress. The "leveling off" of the percentages in the younger speakers is consistent with the typical last stages of a sound change. In a typical sound change, the change begins slowly, picks up speed, and then slows down towards the end.