Assignment
2: Treasure HuntAssignment
2: Treasure HuntThe goal of this assignment is:
This assignment is due on Thursday September 18th, and carries double credit. Answers may be sent by email or handed in in class.
We have prepared a page of potentially helpful resources for you to use. Needless to say, the library is also a mine of useful information on languages and linguistics. However, if you find something useful, please don't take it out of the library: that is grossly unfair to your classmates.
Note: in writing your answers, it is important to briefly explain how you found the answer (e.g. in textbook, in a dictionary, at the Sounds of the World's Animals web site).
Languages of the World
1. BigBurger Inc. is mounting a world-wide advertising campaign to promote its new "Big 'n Beefy" sandwich. The company wants the promotion to be understandable to three quarters (75%) of the world's population, and is trying to figure out how many different languages the advertising must be in to reach this audience. How many languages will they need to advertise in in order to reach this audience?
[Note: for purposes of this exercise, assume that the world's population is 5 billion; also, you can assume that the BigBurger Inc. executives are unaware of the fact that the Big 'n Beefy sandwich is unlikely to appeal to many cultures where vegetarianism is standard.]
2. The country of Papua New Guinea (part of a large island north of Australia) is remarkable for the vast number of languages spoken. If the population of the state of Delaware was so linguistically diverse, how many languages would be spoken here?
[Show how you figured this out: e.g. population of Papua New Guinea, population of Delaware, number of languages spoken in Papua New Guinea etc.]
3. Animals make pretty much the same sounds in whatever part of the world you find them, e.g. if you move my cat to Mexico he would still meow. But are the words that different languages use for animal sounds generally the same? To find out, look for the names that three different languages give to three different animal sounds. Would you say that these names for sounds are arbitrary or are they onomatopoeic? Or are they somewhere in between?
Languages of the United States
4. What are the five most widely spoken languages in the United States today?
5. The vast majority of speakers of Native American languages live west of the Mississippi. Find a Native American language spoken east of the Mississippi river. Where is it spoken?
6. American Sign Language (ASL) is the main language of the deaf community in the United States. It is not a translation of English into gestures, and has its own characteristic phonology and syntax. For historical reasons, it turns out to be quite different from British Sign Language and very similar to French Sign Language. Around how many speakers of ASL are there in the USA today? Can you find a city where there is a particularly large number of ASL speakers (there are a few such cities around the country)?
History and Classification
7. All of these words are borrowings into English from other languages - which languages?
8. Here are some sentences from the Jedi Master Yoda in the film "The Empire Strikes Back" (1983).
What is the basic word order of Yoda's sentences (e.g. SVO (subject-verb-object), SOV etc.). Is this a common word order in languages of the world? If so, can you find a language with basic word order like Yoda's?
Language in Computers
9. Find a word processing program that includes a grammar checking tool (Microsoft Word and Word Perfect are two examples). Try to construct some sentences and have them checked by the grammar checking tool. Is the grammar checker best at picking out (i) violations of prescriptive rules, (ii) violations of descriptive rules, (iii) complete nonsense (word salad)? State briefly how you think the grammar checker is working.
10. [Note: for this question you will need to use a computer which can play sounds that it downloads from the web: all Macintoshes at the computer sites should be able to do this, but many of the other computers will not be able to do this.] One application of linguistics in industry is in the development of computer systems that can take text and turn it into spoken language: such systems are called, not surprisingly, text-to-speech (TTS) systems. AT&T research labs is developing a state-of-the-art TTS system called Voices, which it is letting the public try out via a web site. Creating accurate TTS systems is remarkably difficult, and it is interesting to see how easy it is to fool the system.
One difficulty that a TTS system has to handle is that some strings of letters can be either a verb or a noun (these are called homographs). This is no big deal for words like 'duck', which are pronounced the same whether it is a noun or a verb, but some homographs are pronounced differently, depending on whether they are nouns or verbs. For example, the words 'present' and 'record' have two different stress patterns. The capital letters below indicate a stressed vowel.
To see whether the Voices system can deal with this problem, try out the sentences below.
When does the system need to pronounce present/record as a noun or as a verb? How well does the system succeed at this? Based on when it does and does not succeed, can you get an idea of how the system is deciding whether a word is a noun or a verb?
[Note: don't worry about the fact that the voice sounds rather unnatural, just worry about how the system performs in making the stress appropriate for the noun or verb version of present/record. The Voices demonstration gives you a choice of voices (male, female, child etc.) -- we recommend the male or the female voice, as they seem to give the best quality. On Macintoshes, all of the sound formats offers (au, aiff, wav) should work.]
This is entirely optional, but once you've seen what the AT&T system can do, try out the Model Talker text-to-speech system developed at the University of Delaware's Applied Science and Engineering Laboratory. Does it do any better than the AT&T system?