What is Computational Linguistics?

IPA

IPA stands for International Phoneme Alphabet, which is used in computational linguistics as well as other areas of linguistics to express all the phonemes utterable by the human articulators. The idea is that, although different languages contain different sounds that do not exist in other languages and vice versa, all phonemes can be traced back to one of about 40-45 core phonemes. These core phonemes are what make up the IPA. If a sound does not match perfectly with one of the core phonemes, it is still a variation of one or the result of combining two or more phonemes and pronouncing them simultaneously.

Computational LinguisticsThe IPA is expressed on an IPA vocal chart which is shaped as a trapezoid where small base is south and long base is north. It is shaped such to represent the human articulating cavity. The position of your articulators when you make a certain sound determines at which point on the chart the phoneme is placed. The different points are usually expressed by using terms such as, “open-open-front-back”. Terms like this can pinpoint exactly where on the IPA chart the phoneme falls and thus, exactly how a person must shape his articulators in order to produce the sound. There are actually two different IPA charts: one for vowels and one for consonants. The Vowel chart is a common tool among linguists. The consonant chart is different and I use it as a tool in computation linguistics to calculate vocal equations.

Computational Linguinstics

The point of computational linguistics is to learn how to improve the use of natural language processing by computers and other technology. In order to do this, we have to have a clear understanding of how language is produced and understood. One of the obstacles encountered in CompLing is how to get a computer, a thing that processes information in an entirely systematic and mathematical way, to understand and keep up with the ever changing chaos that is combinatorial grammar.  However, there are algorithms that help the computer pick up on these things.

In order to teach a computer to process information, we have to understand that information ourselves. And since a computer is best at processing information in numerical form, we can turn human speech into a mathematical equation for the computer to read. Note however, that this does not necessarily handle any aspects of semantical data. This method only deals with articulation and sound.

Here’s how it works…

You have an articulator cavity and a set of articulators. The articulator cavity is the empty space of your mouth and is shaped like a trapezoid. Your articulators are all and any part of your body that is used to produce sound. That includes your teeth, tongue, lips, epiglottis, etc… How you position and move your articulators, combined with other variables such as aspiration and stricture, is what makes different sounds.

Now imagine if you wanted to know exactly where and how a sound is produced. One way to do that with high precision is to write a mathematical equation that maps out the positioning and movement of the articulators as the person is talking. This not only teaches you how to pronounce that sound, but it also helps you study the speaker’s own way of articulating because the equation will also tell you about the individual shape and size of his or her articulators.

Every individual has their own voice that is as unique to them as their fingerprint. This means that if two people speak the exact same sentence, their voices will result in two completely different vocal equations. Additionally, the same word might not even be pronounced the same way by the same person two times in a row. A vocal equation explains exactly what was different between the two examples. Computational linguistics helps in navigating these differences.

Because there are so many components that make a voice sound the way it does, vocal equations tend to be very long. It may take a large amount of space to notate a simple document. This is not a new way of communication. It is not something you do to leave notes to yourself or write a letter. It is a way to break down what we hear and study it. However, if you are learning a new language, writing out how to pronounce a new sound or word using a vocal equation is very helping in mastering it.