Applying Phonetics. Murray J. Munro

Code for Information Interchange ) symbols, and are therefore very easy to type on a standard keyboard. SAMPA and ARPABET have fewer symbols than IPA, so they aren't suitable for narrow transcription. However, they are quite satisfactory for computer TEXT-TO-SPEECH ( TTS ) software, which converts text documents to synthetic speech. A critical step in TTS is the conversion of standard orthography into transcription. The process yields an unambiguous representation that can be used to generate a correctly‐pronounced synthetic utterance.

      More details about TTS are presented in Chapter 9.

Table 3.2 compares IPA, ARPABET, SAMPA, and Merriam‐Webster Online symbols for a small selection of English sounds. As can be determined from the table, this is transcribed in IPA as /ðɪs/, in ARPABET as /DHIHS/, in SAMPA as /DIs/, and in the Merriam‐Webster system as \this\.

Table 3.2 Comparison of Several Transcription Systems

Standard Spelling	IPA	ARPABET	SAMPA	Merriam‐Webster Online
top	/t/	T	t	\t\
sit	/s/	S	s	\s\
chin	/tʃ/	CH	tS	\ch\
these	/ð/	DH	D	\th\
seat	/i/	IY	i:	\ē\
fit	/ɪ/	IH	I	\i\
cat	/æ/	AE	{	\a\

      for further thought, analysis, and discussion

      1 Considering only the initial sounds in the words below, identify the odd‐item‐out in each triplet.a) piefivephoneb) threethinkthatc) sizeshoppsychod) hugehousehoure) sheepsugarsuchf) keepknowneedg) anyedgeaskh) useupunusuali) hourownowlj) underunionyield

      2 Considering only the final sounds in the words below, identify the odd‐item‐out in each triplet.a) tombrobsomeb) coughlaughthoughc) sayregretcrochetd) spicegazetriese) workedputrodef) trayhighdryg) dicemouserouseh) songthingbigi) donosoj) gaugerougefudge

      3 Provide one or more homophones for each of the following.a) pairf) keyb) teaseg) throughc) ourh) nosed) alli) packede) seesj) tax

      4 Using only the symbols in Table 3.2, “translate” the following words into each transcription system:WordIPAARPABETSAMPAMW‐Onlinea) teachb) thatc) east

      Suggested answers for items 1–4 are available on the APSSEL website.

for further reading

      If you are curious about the idiosyncrasies of English spelling and their origins, you might enjoy Upward and Davidson's The History of English Spelling (2011). See the List of Sources at the end of the book for publication details.

4 The Sounds of Language

      One of the main goals of phonetics is to describe how speech segments are produced using the structures of the vocal tract. The sounds of the world's languages can be characterized—at least in a general way—using a relatively small number of descriptive labels. As a starting point, we need to recognize that all languages have sounds falling into the two major categories of SEGMENTS: CONSONANTS and VOWELS. In this chapter, we will survey the inventory of sounds used by speakers of English, with a few digressions along the way to compare regional English varieties. We'll briefly consider some sounds used in other languages as well.

4.1 consonants

      Consonant sounds are produced by obstructing the airflow through the vocal tract—either completely or partially. To describe consonants, phoneticians use three main dimensions:

      1 Manner of articulation : the degree of the obstruction.

      2 Place of articulation : the location of the obstruction.

      3 Voicing : the state of the vocal folds (vibrating or not).

      The first two relate specifically to ARTICULATION: the act of configuring vocal‐tract structures to manipulate airflow. The third, VOICING, has to do with PHONATION, which refers to the use of the vocal folds to generate sound.

      4.1.1 manner

When discussing the manner of articulation of consonants, it's helpful to think in terms of how much oral obstruction is involved: complete, partial, or very slight. Table 4.1 summarizes the possibilities for English. In the case of plosives (also called oral stops), the obstruction is complete: when you say pay, you produce the plosive /p/ by first closing your lips completely to block the oral airflow and then opening your lips to create a very short explosion of air. Notice that you hear nothing at all during the closure; it's only when you release the stop that any sound is produced. The blockage at the lips is the same for the nasal consonant /m/; however, during the /m/, you allow air to pass through the nasal cavity, and sound occurs throughout the oral closure. While it is possible to sustain the /m/ sound for quite a long time—until you run out of air—the sound of /p/ can't be sustained since it occurs instantaneously on the release. Other plosives include /b/ as in boy and /t/ as in tot, and other nasals are /n/ as in tot and the /ŋ/ sound at the end of baug .

      Partial obstruction of the vocal tract occurs for fricatives , such as /f/, which you produce by bringing your lower lip and upper teeth in contact while allowing some air to pass through the small opening between them. The result is a noisy sound that can also be sustained for some time. Other fricatives in English include /s/ as in sank and the /ʃ/ at the beginning of show.

      A third manner is affricate. The ch combination in chin represents an affricate consonant in which a complete air blockage is followed by a partial one. This means that an affricate is actually a plosive followed by a fricative, and ch is represented by the IPA symbol /tʃ/, consisting of /t/ as in tot and /ʃ/ as in show. Even though this consonant is transcribed with two symbols, it functions as a single sound