Date of Award


Degree Type


Degree Name

Doctor of Education



Major Professor

Paul C. Burns

Committee Members

J, Estill Alexander, Lester Knight, John Peters


The purposes of this study were to analyze the pronunciation of pupils and teachers to determine the deviation from the expected pronunciation for each group, and to determine the influence of dialect on reading performance of pupils. A random selection of twenty first grade, third-generation pupils and twenty elementary, second-generation teachers from four small, rural, all-white elementary schools in Sevier County, Tennessee were interviewed with the use of an instrument which the investigator developed from an arbitrary selection of one-hundred-sixty-two pictorial words. These words were teaching examples for the eighteen vowels and diphthongs that are taught as phoneme-grapheme correspondences in three leading basal reading programs.

Each interview was conducted individually and taped on a reel-to-reel recorder. An informant (pupil or teacher) had only to look at a picture and say what it was.

The material was transcribed from the tapes using the revised International Phonetic Alphabet and later compared with the teaching examples taken from the three basal reading programs. Any deviation from the expected pronunciation was counted as a "divergence" and the score for each informant became the number of such divergences.

The divergence scores for both pupils and teachers were analyzed statistically, and the results indicated that neither the pupils nor the teachers pronounced the teaching examples according to the expected pronunciations to which they were compared, and that there is a difference between the speech of the pupils and the teachers. While the teachers tend to more closely approach the expected pronunciations, there is as much variation among teachers as among pupils.

The pupils were administered a standard reading test (vocabulary part), and these raw scores compared to their individual divergence scores. The results showed a correlation coefficient of only -0.29. The lack of significant correlation between pupils' pronunciation and reading scores is taken to indicate that dialect differences in their phonological aspect are not related to reading performance.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."