Real Life Phonology

Linguistics is a relatively young science, so it’s still got a few kinks to work out. One I just recently noticed is that the definition of “phoneme” needs a little work. The official definition is a little hard to pin down, but let’s go with the great Wiki’s “the smallest segmental unit of sound employed to form meaningful contrasts between utterances.” That’s a good enough definition for most purposes. However, phonemes are referred to by their base allophones, so I suppose that’s really what I have a problem with.

See, you don’t say a sound the exact same way every time. Phonemes are a way of helping us to deal with that (our brains only need two instructions to build words: 1) string together the proper sounds, 2) change sounds appropriately to fit surrounding sounds; thus, “skill” and “kills” have the same sounds in different orders rather than completely different sounds… sorta). The different sounds that can all be attributed to the same phoneme are called allophones; basically, a phoneme is an abstract concept of a sound, whereas allophones are the actual sounds that come out (if I weren’t writing in English, I might even be tempted to draw a comparison to writing vs. speaking). Linguists don’t specify phonemes by a single sound, however; they will use either a /broad transcription/, in which they use sounds they personally think of as phonemes of the example language for this particular purpose, or they’ll use [narrow transcription], which uses the allophones that a speaker actually uses, and thus has no use for phonemes.

Got it? If not, check out The Language Instinct by Stephen Pinker, a great layman’s introduction to linguistics.

Anyway, my problem with this definition and use of the concept of a phoneme is that linguists only think of phonemes as groups of allophones, whereas native speakers actually do choose an allophone to use as the base phoneme. How native speakers of a language think is not only a legitimate basis for research, it is the basis for a lot of the rules we have today, and even for the existence of phonemes in the first place.

I suspect that the problem linguists have with phonemes is the difference between how linguists come up with a base allophone (for use in broad transcriptions) and how native speakers come up with a base allophone. See, most allophones occur in predictable environments such as at the beginning of a word (#_), at the end of a word (_#), after a palatal consonant ([+pal]_), etc. The base allophone is the one that is the least predictable (generally, the one that can be generated with the rule “in all other contexts”). Native speakers, however, choose the one that they consider “strongest,” which is an entirely personal choice… and yet, one that will generally be the same for all people who speak a given dialect.

Here’s an example: a linguist studying the English language will notice that there are two main allophones for l: [l] (an alveolar lateral approximant, commonly called “light l”) and [ɫ] (a velarized alveolar lateral approximant, commonly called “dark l”). This linguist will further notice that [ɫ] only occurs at the ends of words (“deal” [di:ɫ]), whereas [l] occurs in various contexts (“litter” [lɪɾɹ̩], “slip” [slɪp̚]). (Further note that l is not at the beginning of the word “slip.” It is in the first half of the syllable, yes, but in phonology you have to remember that it is preceeded here by the s.) Therefore, the linguist will conclude, [l] is the base form of the phoneme, and thus he will use /l/ in broad transcriptions. However, if you were ever to ask a native English speaker (especially a North American one) what sound he or she thinks best represents what is represented in the above three words with the letter “l,” he or she will respond [ɫ]. Pretty much all the time.

Another example: In American English, all voiceless plosives(/p, k, t/) have three main realizations: aspirated syllable initial, followed by a vowel ([pʰ]/σ_V), unreleased syllable final ([p̚]/_σ), and unmodified elsewhere ([p]/elsewhere). Examples: “pail” [pʰe͡jɫ], “stop” [stăp̚], “spill” [spɪɫ], “play” [ple͡j]. The linguist will now conclude that [p] is the base form of the phoneme, whereas if you ask a native English speaker to give the base sound for this phoneme they will respond with the aspirated form.

This is not to say that these two ways of finding the base form of a phoneme never match up. For example in Modern Standard Arabic the sound represented in spelling by a fatha has allophones [a, æ, ɛ], of which [a] is recognized by both native speakers and linguists as the base form. (If you’re an Arabic speaker who disagrees, chances are you’re one of the 99% of Arabic speakers who doesn’t speak MSA. Unfortunately, I’ve only taken one and a half Arabic classes, so that’s all I’ve had experience in.) However, they don’t always line up, and I haven’t seen anything that implies linguists care. I realize that to some extent linguists have to streamline the rules they’re working with because language is too complicated for any computer currently extant to handle (there’s a reason people are abandoning generative grammar as anything more than an introduction and theoretical exercise), but they do at least tend to address the differences between theory and reality. When one particularly annoying student constantly brought up things like differences between the word “either” and the logical definition of “either,” my semantics professor in college used to tell us, “You are not writing English sentences. You are writing Semantic Logic sentences that usually resemble English.” However, that difference is never addressed in Phonology, and it just annoys the heck out of me.


0 Responses to “Real Life Phonology”

  1. Leave a Comment

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

July 2010
« Apr   Aug »

%d bloggers like this: