Lab 4 Syllable Restoration: Finding materials

The first part of the lab involves finding the words and creating the nonwords to use. For this part of the lab, each person will be assigned several letters of the alphabet. You are responsible for 1) using the MRC Psycholinguistic Database to find all the words which fit the below constraints that start with whatever letter you were assigned; and 2) creating 20 nonwords (beginning with any letter) that fit the below constraints.

Your list of words is due Monday April 7^th by noon. Since we’re on a fixed schedule, we can’t accept any late work—we will start compiling the lists on Monday afternoon so that we can have them ready for recording by Wednesday.

1. Word items

Constraints:

- must be morphologically simple

- must be 3 syllables long

- should not be super-rare

- should begin with a consonant

- ideally, should be fully identifiable as the word based on any 2 of the 3 syllables

o e.g., bachelor – __chelor … ba__lor … bache__

To find the words, go to http://www.psy.uwa.edu.au/mrcdatabase/uwa_mrc.htm. In the first step, select ‘Word’, ‘Number of letters’, ‘Number of syllables’ and ‘Kucera-Francis written freq.’ In the second step, set upper and lower limits for ‘Number of syllables’ as 3 and 3 (see picture below); this will let you output only the 3-syllable words. Also select a lower limit of ‘1’ for Kucera-Francis frequency; this will make sure you don’t get many misspellings or archaic words.

Finally, you should filter so that you only get words that start with your selected letters. If you were assigned ‘B’ and ‘C’, under ‘SIMPLE LETTER MATCH’ you would enter the pattern string [bc]*. The star indicates that anything can come after the first letter.

Now go down to the bottom and click the ‘GO’ button. If everything worked correctly, you should see a long list of words that fit the constraints you specified. Now you should copy and paste them into some kind of file (*paste the whole line, including the frequency info etc), and then go through and eliminate any items which fail for another reason—mainly if they seem extremely rare or weird to you, or if they are morphologically complex. Words with any kind of transparent prefix or suffix like ‘-tion’ or ‘-er’, etc., are out.

You should send the remaining good candidates to us by email. Please try to use a common document format!

2. Nonword items:

- must be 3 syllables long

- should obey phonotactics of English (e.g. not something like tlop)

- should start with a consonant

- should not be possible words if one of the syllables is removed

o e.g. batelor would be a bad choice because when the second syllable is covered in noise, ba__lor, it could be bachelor

You have much more freedom in creating the nonword items—you can make up whatever kind you want as long as they fit the constraints above, and can start with whatever letter you want. An easy way to make sure that they obey the phonotactics of English is to just make sure that they are easily pronounceable. If you have trouble just coming up with words out of the blue, you can take existing words and alter phonemes in each syllable until you come up with a completely different nonword. Just make sure that the nonwords you come up with are not too similar to any real word. In particular, as mentioned above, make sure that they do not make possible words when one syllable is covered in noise: the whole point of the experiment is to test how well people can distinguish between words and nonwords in noise, so we can’t have nonword stimuli that are possible words when noise is covering parts of them.

Again, send the list of 20 nonwords to us by email. If you get done early, please go ahead and send them immediately, that will help us get a head start on compiling everyone’s words!

As you’re doing this task, think about whether there are other constraints that might be good to implement, if we have enough item candidates—one that was mentioned in class was stress.

Alphabetic assignments

Shannon – [B C]

Janet – [D F G]

Emma – [H L M]

Adam – [N P]

Jen – [Q R]

Hana – [J S]

Tara – [K T]

Danielle – [V W X Y Z]

Amanda will be recording our materials, so you will get to hear a lot of her voice in the next couple weeks!