Second languages, dead languages, AI, poetry, etc.

These readings gave me a deeper understanding of the types of sounds in speech,and how those sounds are produced. The Gusshoven article informed me about the anatomy and physiology of the organs of speech. I learned that the larynx produces either no sounds - called “voiceless” - or phonated sounds, which can be normal, breathy, or creaky. The Kenstowicz article addressed some of the challenges of phonetic representation. Single underlying sounds - phonemes - can have many different variations - allophones - that speakers don't distinguish. Phonologists try to determine the rules that dictate which allophone a speaker will pronounce in a given word; Kenstowicz names some of the challenges of and varying theories about finding these rules.

Kenstowicz points out that speakers often don't distinguish differing pronunciations of the same phonemes. I've certainly experienced this. Since I came to Stanford, my friends have joked about the way I pronounce “o” - to them my “econ” sounds like an exaggerated “ecahn” and my “job” is “jahb.” I never noticed a difference. Even though I didn't notice this difference in my own speech, it's interesting to think about how much of my own observation of others' speech could be subconscious. Do we notice subtle differences in the way other people speak and draw conclusions about them?

Throughout the reading, I kept thinking that my understanding of linguistics is almost certainly impaired by the fact that I only speak one language, English. For example, the Gusshoven article pointed out that English only has pulmonary egressive sounds - sounds created with help from the lungs that are pronounced in exhalation rather than inhalation. This observation is almost meaningless to me since I haven't spoken words in other languages that don't follow this pattern. It still seems possible to study and categorize phonetics in different languages without actually speaking them, but it does seem that understanding more languages would make you a better linguist. Particularly if, according to Chomsky's hypothesis, phonetic rules of a language represent hypotheses about the way information is stored in the brain, it would seem difficult to fully understand these phenomena without personally experiencing them. Unfortunately in high school I studied Latin, and understandably, we didn't study Latin speech.

(As a side note, the Kenstowicz reading did leave left me wondering if there exists any academic research on ancient phonology. What are the methods for trying to determine if “Cicero” was pronounced “Sissero” or “Kickero”? Did ancient phonemes have more or fewer possible allophones than today's English phonemes?)

Throughout this reading assignment I thought about the connections to my other reading and classes about philosophy and artificial intelligence. It was surprising to me that we can articulate different sounds so instantaneously, using remarkably variable anatomical mechanisms, without conscious thought. It seems very different from learning to run or ride a bike - these are large, relatively simple motions that we make a conscious mental effort to learn. Although I don't know much about it, it seems like language acquisition requires less will. We learn the sounds of certain letters, but we don't actually need to pay attention to how we operate our organs of speech to produce those sounds.

This reminded me of some of Descartes' observations that led him to think that the mind and body were separate - he observed actions that seemed to be purely reflexive, like jerking your knee when someone hits it. Production of speech seems to fall into this reflexive category - while we consciously decide to enunciate a word, we don't consciously think about moving our mouths and other organs of speech in the right way to produce the sounds. Descartes concluded that since some of our actions were unconscious, with purely biological causes, while others were conscious, the result of mental decision, our mind was separate from our body. Of course this most likely isn't true, but it's still interesting to consider how other theories of mind and body, such as functionalism or the identity theory, explain linguistic function, or how understanding how speakers distinguish and identify phonemes can assist in creating artificial intelligence.

I was unsure about one of Gusshoven's points. He seemed to imply that English only has left-dominant feet. A foot is usually a two-syllable rhythmic unit; a left-dominant foot will have emphasis on its second syllable. If this is what he saying, it doesn't seem accurate: there are plenty of words that violate this rule: “produced,” “occur,” and others. It is interesting that in poetry, many of the natural stresses in a word are distorted to fit into meters like iambic (unaccented/accented - if both syllables were in one word this would be left-dominant), trochaic (accented/unaccented - if both syllables were in a word this would be right-dominant) dactylic (accented/unaccented/unaccented), or anapestic (unaccented/unaccented/accented). It would be interesting to study how phonemes are produced abnormally from normal speech in poetry.