Variation across part-of-speech for talk

To get some practice with POS-tagging, we will take a brief look at the word talk. We will compare the frequency of the noun talk(s) with the verb talk(s) according to two categories: written informative texts, and spoken demographic (i.e. conversational) texts.

Before you start, follow this link and make a copy of the table - either print it out or copy-and-paste it over to a word processor. Then you can use it to keep a note of your results.

You might wish to share this task with another student, with one of you doing the spoken and one of you doing the written.

For both spoken and written texts, follow the steps listed below:

1. Start WordSmith
2. When choosing texts, double click into the folder BNCsamp, then into taghoriz
3. Now move to either spoken or the folder
4. Then
   1. If you have chosen spoken select the folder demog, pick all the files, then start the Concord program
   2. If you have chosen written select the folder inform, pick all the files, then start the Concord program
5. Search for the following to find talk as a singular noun:
   

   talk_NN1

6. Fill in the frequency for the appropriate section into the results table
7. Now look for talk as a plural noun (use the tagset if unsure) and fill in that frequency in the table
8. Now look for talk as a verb. There are two tags available: VV0 and VVI. The best way to look for them in one search is to use slash ( / ) meaning "or"

   talk_VV0/talk_ VVI

9. Repeat for talk as a third person singular verb, adding the score to the table
10. If you are working with a partner, get their results to complete the table. If not, repeat the steps above for the text-type you didn't do the first time.

DON'T WORRY ABOUT THE TABLE'S "PER MILLION" COLUMN FOR NOW.