Entry #3: Phonology, or The Sounds of Silence

While Phonetics (and phoneticians) seeks to describe the physical sounds of language, Phonology (and phonologists) seeks to categorize and explain the sound patterns of a language. What sounds is the human vocal tract capable of producing? Why do some languages use certain sounds but not others? How concrete is the sound inventory of a language? How does a language use its inventory to convey differences in meaning? The phonologist, through data collection and analysis, seeks to answer these questions.

Let us take English as an example. While a phonetician would study the way a speaker produces a /ŋ/, like the average length, organs used, airflow, etc, a phonologist would study why and where this sound is produced. Why is it only found at the ends of syllables, such as in "hang" or "string?" Is it only found at the end of syllables? Do some dialects/speakers pronounce it elsewhere? Yes and No, respectively. Mountains of data confirm this. The more interesting (and more studied) question is, why is this so?

How can we reconcile this fact with the fact that Kabiye, a Niger-Congo language spoken in West Africa, easily places this sound at the beginning, middle, and end of syllables? Clearly, the restriction in English does not arise from a physiological standpoint: the human vocal tract is quite capable of pronouncing this sound in many different environments. Kabiye is not alone; many languages have this feature.

The solution lies in parameters, or DNA-encoded instructions for how one is to speak. There is a little switch in our heads (in fact, there are many) that get flipped when we first hear language around us, usually 0-2 years old. A child listening to English would never hear /ŋ/ at the beginning of syllables; a certain switch is flipped to "no," and thereafter it is very hard to change that. Sometimes, even trained phoneticians have trouble pronouncing non-native sounds. A baby growing up where Kabiye is the most frequently spoken language, however, would have their switch flipped to "yes." Countless numbers of these switches can account for the tremendous amount of linguistic diversity found on earth.

Entry #2: "read read lead lead" or /rid rɛd lid lɛd/

As early as the 20th century, linguists decided that they needed a uniform code with which to transcribe the world's languages. English spelling didn't work, because it is not uniform or patterned in any way. In order to have a cooperative, uniform, scientific process, some particularly inventive phoneticians came up with the International Phonetic Alphabet.

The IPA, as it is known, is a written code made up of mostly Latin characters, with a few Greek ones thrown in for kicks. Each symbol represents a sound. Each transcription is written with slashes or brackets around it, to differentiate the transcription from other written communication. For example, the English sentence "my name is Linguo, hear me roar" is transcribed as /maɪ nejm ɪz lɪŋgwow hir mi rɔr/. Notice a few points of interest: 1. each sound has its own symbol; 2. in this example, the number of characters needed does not differ terribly from English, but this is not always the case. "knowledge" is /nalɪʤ/; 3. phonetic transcription looks English-like, but is definitely not English.

Transcribing utterances (meaningful speech sounds) with a uniform alphabet is a fundamental part of Linguistics. In an effort to describe the world's languages as closely as possible, it is probably foremost important that we see which languages have which sounds, and which do not. For example, the sound /ɸ/ is absent from English, but not Spanish (it begins the words vaca:cow and vez:iteration). As you can see, if we used a nonuniform system, these two sounds would be identically written, and thus treated the same.

There are disadvantages to phonetic transcription, however. It represents speech as disparate, like a sequence of independent sounds. Analysis using computer programs (like Praat) has shown us that speech is a continuous, flowing phenomenon; there is no silence when one speaks (other than deliberate pauses). Transcription ignores this fact, and abstracts to an approximation of the different sounds heard. Indeed, hearing is the operative word: ought we to transcribe what I hear, what you hear, or what the speaker thinks he is saying? Even with sophisticated computer programs, experienced phoneticians still disagree about just what sound they heard.

Resources:

1. A clickable audio representation of all the IPA sounds http://www.ling.hf.ntnu.no/ipa/full/ipachart_cons_pulm_fbmp3.html

2. official IPA website:
http://www2.arts.gla.ac.uk/IPA/ipa.html

3. a full-size chart of the sounds of the world's languages: