Applying Phonetics. Murray J. Munro
kinds of drawbacks, as shown in Figure 3.1. First, it has numerous one‐to‐many relationships, in which a single letter can correspond to multiple phones. For instance, the letter “a” can represent the different vowel sounds in cat (/æ/), Kate (/eɪ/), and any (/ɛ/). Second, it also has numerous many‐to‐one relationships, where a particular phone is spelled in more than one way. For example, the first sound in she, /ʃ/, is spelled with “sh,” but the same sound in sure is written with “s”, and in nation is written with “t.” English spelling therefore lacks the precision needed for a scientific tool.
2 They can represent the sounds of languages that have no orthography.When European anthropologists first interacted with the Indigenous peoples of the Americas, they needed a phonetic alphabet in order to document scores of unwritten languages. That need still exists because languages previously unknown to scholars continue to be identified in the twenty‐first century, such as Pirahã (Brazil), Koro (India), and Light Warlpiri (Australia)—all discovered within the last decade or so.
3 They can represent the sounds of languages that have a written MODALITY but do not use an alphabetic orthography.A number of languages, including the many varieties of Chinese, have writing systems that generally do not indicate pronunciation. A phonetic alphabet allows us to represent the sounds of these languages systematically and unambiguously.
4 They can represent sounds in the same way for all languages.The pronunciation of the Spanish word sí, meaning yes, is very close to that of English see. Notice, however, the difference in the spelling of the vowel sound. By using a single phonetic symbol /i/ for both Spanish and English, we can capture the fact that the two languages have nearly the same vowel despite the orthographic difference.
5 They can systematically represent sound distinctions that are not reliably indicated in the standard orthography of alphabetic languages.English spelling does not distinguish the first sound of thin (/θ/) from the first sound of this (/ð/). Though both are spelled th, if you produce them in isolation you will feel and hear a clear difference. This is due to the presence of vocal fold vibration in the /ð/ of this but not in the /θ/ of thin. Using separate phonetic symbols captures this important distinction.
TRY THIS
☛ 1. How many different spellings can you think of for the sound of the vowel in bed (IPA: /ɛ/)? Hint: There are more than five. 2. How many different pronunciations can you think of for the letter “t”?
Answers: 1. At least nine: any, said, says, bed, head, heifer, friend, foetid (British spelling), bury (some speakers). 2. At least six: two (/t/), nation (/ʃ/), listen (none), three (/θ/), these (/ð/), culture (/tʃ/)
An interesting kind of complexity in sound and spelling arises from homonyms (from the Greek for “same name”), words that sound or look the same. The Venn diagram in Figure 3.2 gives a breakdown. Though not all authorities agree on how to define homonym, most recognize two main subtypes: homographs, different words with the same spelling (the two meanings of close); and homophones, different words with the same sound (there and their). At the intersection are words that fit both categories simultaneously, such as lie as in “recline” and “an untruth.”
Figure 3.2 Types of homonyms
3.2 phonetic transcription
Phonetic transcription is an IMPRESSIONISTIC process used by linguists to represent the discrete, recombinable sound units of language in visual form. By this we mean that phoneticians use their ears to identify the sequence of sounds that someone has produced and then record their analyses as a series of written symbols. Impressionistic phonetic transcription can be used in a BROAD way to capture the basic sound categories used by the speaker or in NARROW fashion to specify many of the fine details in speech. Skilled phoneticians can transcribe speech reliably and with a high degree of precision, but training and experience are essential, and even experts don't always agree on how a particular utterance should be transcribed. Sometimes there are debates about the use of particular symbols, and other disagreements arise about terminology. Still another complication has resulted from the typesetting difficulties that phonetic transcription posed in much of the twentieth century. Everyday typewriters did not have symbols like /ʃ/ and /ð/, and even commercial printers were often unable to reproduce phonetically transcribed utterances. As a result, more easily typed symbols were sometimes used, such as /š/for /ʃ/. With contemporary word‐processing applications, these problems have largely disappeared, but many older sources of speech data are still in use. For all these reasons, you will find that textbooks (and instructors) sometimes differ in the particular choices they make when doing transcription. You will also discover that American, Australian, British, and Canadian sources differ from one another because of dialectal variation. While these matters may seem challenging, there is no use in worrying or complaining. Speech is a complex phenomenon that even phoneticians only partially understand. This is a reality with which everyone who works in the field must come to terms. This book uses transcription mainly as a tool for discussion of other concepts. For the most part, we will use broad transcription in which only the major speech sound distinctions are captured, each with a different symbol.
In this context, the term impressionistic does not mean “casual” or “unscientific.” Instead, it refers to “systematic analysis by ear,” rather than with instrumentation. Experienced, highly‐trained phoneticians are capable of perceiving remarkable levels of detail in speech and of capturing what they hear on the printed page.
3.2.1 transcription systems
The IPA is the transcription system most commonly used by phoneticians. You need to know some of the IPA symbols in order to understand many of the issues that arise throughout this book. To distinguish broad IPA transcription from standard orthography, we place the symbols between forward slashes (/ /), also known as virgules. School, for instance, which is spelled with six Roman letters, actually consists of only four sounds and is transcribed in IPA as /skul/—a sequence of four phonetic symbols. In some publications, you will also see narrow transcriptions, which are placed within square brackets ([ ]), and which show more detail than is usually necessary in this book.
TRY THIS
☛ How many discrete sounds can you count in each word below?
a) know; b) nose; c) thing; d) chewed; e) ship; f) box; g) make; h) whose; i) thick; j) thought
Answers: a) 2; b) 3; c) 3; d) 3; e) 3; f) 4; g) 3; h) 3; i) 3; j) 3
Systems other than IPA may also be used for transcribing speech. Printed dictionaries and some web‐based reference materials designed to help readers with word pronunciations use their own sets of symbols, together with a key. For these users, getting the right pronunciation of a new word depends on reading the symbols correctly. It is possible that these dictionary‐based systems will soon become obsolete because of the availability of links to audio recordings and synthetic speech apps that allow the user to hear a spoken production of a word with no need to interpret symbols.
Instead of IPA, speech technologists often prefer to use SAMPA (Speech Assessment Methods Phonetic Alphabet ), ARPABET (Advanced Research Projects Agency + “BET” ), or some modified version of either one. Both systems permit a broad transcription of English words using