University of Arizona, Department of Computer Science

CSc 120: Phonetics and Pronunciation

There are two components to figuring out the pronunciation of an English word: the sequence of sounds involved, and the emphasis (or stress) on different syllables. This document describes how these two components are described in the Carnegie-Mellon Pronunciation Dictionary.

Phonemes

A phoneme is a unit of sound. To describe the sounds uttered when pronouncing a word, we give the sequence of phonemes that entails.

One issue that we have to deal with is that in English, the same letter can be pronounced in a number of different ways, e.g.:

the letter i in the words "inside" and "item";
the letter c in the words "cat" and "certain".

The the CMU pronunciation dictionary gets around this problem by using pairs of letters to indicate specific sounds, i.e., phonemes. There are 39 such phonemes; they are shown here. For example:

the phoneme IY describes the "ee" sound in the words meek, weak, me, and knee;
the phoneme TH describes the "th" sound in sixth and thousand; and
DH describes the sound of the letters "th" in thee and weather.

Stress

In addition to the sounds of a word, different syllables of a word are often emphasized, or stressed, differently. Different patterns of stress can correspond to different word meanings, as this story illustrates:

"A pilot once told me a story about a member of his cabin crew informing him that they had an inVALid passenger on board. The pilot was a bit confused and wondered if the passenger didn’t have a ticket or wasn’t on the roster. He couldn’t figure out what the problem was. It wasn’t until the flight attendant continued to explain that the passenger was in a wheel chair and had special needs that the pilot realized he meant to say INvalid!"

The CMU pronunciation dictionary indicates stressed syllables by adding one of the numbers 0, 1, or 2 to the phoneme for the vowel sound of that syllable, with the following meaning:

0 : no stress
1 : primary stress
2 : secondary stress

For example:

Word Phonemes Stressed syllables

weekly W IY1 K L IY0 The main (primary) stressed syllable is week; this stress is indicated by the '1' in the phoneme IY1.
The syllable "ly" is unstressed; this is indicated by the '0' in the phoneme IY0.

computerize K AH0 M P Y UW1 T ER0 AY2 Z The main (primary) stress is on the syllable put; this is indicated by the '1' in the phoneme UW1.
The secondary stress—i.e., a syllable that has some stress, but not as much as the primary stressed syllable— is ize; this is indicated by the '2' in the phoneme AY2.
The other two syllables in this word ("com" and "ter") are unstressed; this is indicated by the '0' in the corresponding phonemes.

Word	Phonemes	Stressed syllables
weekly	W IY1 K L IY0	The main (primary) stressed syllable is week; this stress is indicated by the '1' in the phoneme IY1. The syllable "ly" is unstressed; this is indicated by the '0' in the phoneme IY0.
computerize	K AH0 M P Y UW1 T ER0 AY2 Z	The main (primary) stress is on the syllable put; this is indicated by the '1' in the phoneme UW1. The secondary stress—i.e., a syllable that has some stress, but not as much as the primary stressed syllable— is ize; this is indicated by the '2' in the phoneme AY2. The other two syllables in this word ("com" and "ter") are unstressed; this is indicated by the '0' in the corresponding phonemes.