Linguistics 001      Fall 2011     Homework 4      Due We 10/12

1. Pitch range comparison

The main point of this exercise is to get you used to using a computer program for acoustic analysis of speech. In order to make the exercise more interesting, you'll evaluate a gibe by Andrew Sullivan directed at Mark Levin and Sarah Palin.

In a blog post on 10/5/2011, Mr. Sullivan discusses Sarah Palin's announce of her decision not to seek the GOP nomination for president in 2012. He introduces his discussion this way:

Palin talks to Mark Levin here (her voice is the deeper one).

Mark Levin is a radio talk show host, and the link goes to a page on his web site that includes not only the text of Palin's statement, but also accesses this mp3 file of a 15-minute segment of his show. (For your convenience, I've put a .wav version sampled at 11025 Hz here.)

Choose a convenient folder to work in, and download the interview's audio from one of the sources given above. (In Firefox on Windows machines, use right-click>>Save link as... With other operating systems or other browsers, use the appropriate corresponding method. If you can't figure it out, ask your TA or your ITA.)

Now download one of these two: the free program Wavesurfer, or the free program Praat, which is a bit harder to learn to use, but can do many more things. In your recitation section, you'll learn a bit about how tto use these programs to analyze audio recordings.

Your task here is to evaluate Sullivan's claim -- is Sarah Palin's voice actually "deeper" (interpreted as "lower in pitch") than Mark Levin's voice?

There are several ways to approach this. For example, you might select a few phrases from each speaker, and measure the peak pitch in Hz (cycles per second) for each word that contains a significant local masimum in pitch. If we do this for a phrase produced by Mr. Levin (reading from Ms. Palin's prepared statement) starting at around 84.8 seconds into the file, we get something like

Know [177] that by working[187] together [184]
we can bring this[156] country[136] back[185]

Having done this for what you believe to be a representative sample of each speaker, you could then compare the distribution of numbers that you've found.

A more ambitious alternative would be to get all the F0 (i.e. pitch) values from your selected phrases -- in Praat you can use the Pitch>>Pitch listing menu item -- and do your analysis on them, perhaps along the lines suggested here.

2. Telugu morphology
(based on an exercise in Farmer & Demers' A Linguistics Workbook):

The table below gives a set of 22 words in Telugu, each of which is translated by an English sentence. (This is an uncharacteristically simple sample of Telugu verb morphology.)

English translation
1 ceppɛɛnu "I told"
2 ceppincunu "I cause (someone) to tell"
3 cuustaam "we will see"
4 ceppɛɛm "we told"
5 ceppanu "I will not tell"
6 navvincum "We cause (someone) to laugh"
7 cuustunnaaḍu    "He is seeing"
8 ceppɛɛyi "They told"
9 koḍataanu "I will beat"
10 paaḍataanu "I will sing"
11 ceppɛɛru "You (pl.) told"
12 ceppavu "You (sg.) will not tell"
13 ceppɛɛvu "You (sg.) told"
14 ceppam "We will not tell"
15 ceppɛɛḍu "He told"
16 cuusɛɛḍu "He saw"
17 cepparu "You (pl.) will not tell"
18 koḍatunnaayi "They are beating"
19 ceestunnaanu "I am doing"
20 aḍugutaam "We will ask"
21 ceesɛɛnu "I did"
22 aḍugutaaḍu "He will ask"

A. List the Telugu morphemes corresponding to these English words:

  Telugu morpheme English gloss
1.a          "tell"
1.b   "sing"
1.c   "see"
1.d   "laugh"
1.e   "ask"
1.f   "beat"
1.g   "do"
Person marking of subjects:
2.a   "I"
2.b   "you (sg.)"
2.c   "he"
2.d   "we"
2.e   "you (pl.)"
2.f   "they"
3.a   past tense
3.b   present tense
(-ing form in English gloss)
3.c   future tense
3.d   negative future tense
3.e   causative

B. List the order in which the morphemes occur in the Telugu words. Use terms such as verb, tense, and subject.

    [course home page]    [lecture schedule]     [homework]