Texts, Tweets, Online Forums and Speech

Multi-modal talk

Texts, tweets, online forums and speech

Before embarking on your project, please bear in mind that there are ethical issues involved in collecting text messages and recording/transcribing conversations. To be respectful and ethical in your approach, you should obtain permission from individuals before collecting and using their data. If somebody doesn’t give you permission, don’t use their data – find another source.

Data collection

For this project you will need to collect a number of samples of different types of language use. You will then be analysing your data.

Text messages

Collect twenty to thirty text messages in a word processing file. (If you can download them directly, that is fine, but otherwise you will need to re-type them.)

1. Make sure they are not all written by you or to you. Share messages with your friends.

2. Very short messages (those of only one or two words) will be much harder to work with than longer ones, so keep the number of very short ones to a minimum.

3. Change names to letters such as ‘A’, ‘B’ or ‘C’, or to made up names, so nobody can be identified.

4. Substitute ‘*?*?’ for any obscenities (swear words).

5. Don’t include any messages that could offend or upset someone else.

Online forums

Visit several online forums. Below is a list of possibilities, but you don’t have to use these:

www.pcadvisor.co.uk/forums

http://community.manutd.com/forums

www.makeup-forum.com

www.gamesforum.com

www.muzicforums.com

Copy ten to fifteen posts. Try to pick them from at least four different sites.

Tweets

Collect about thirty tweets from http://twitter.com (or another microblogging site). Make sure they are not all from the same person and they are on different topics. Be a little selective. In particular, don’t use tweets that are just copies of something else.

Spoken language

Record you and your friends or family talking on a number of different occasions. Transcribe about five minutes of speech. What you transcribe doesn’t have to come from just one conversation.

Analysis

Tweets, text and online language

Before you start your analysis of the data you have collected, consider the following about online and text language.

1. Single letters can replace words, for example, ‘c’ for ‘see’, or ‘r’ for ‘are’.

2. Single digits can replace words, for example, ‘4’ for ‘for’.

3. A single letter or digit can replace a syllable, for example, ‘l8r’ for ‘later’.

4. Alternate spellings are used, for example, ‘no’ for ‘know’.

5. Phrases can be represented by the first letter of each word in the phrases, for example, ‘ttyl’ for ‘talk to you later’, or ‘btw’ for ‘by the way’.

6. Words can be represented by pictograms, such as ‘<3’ for ‘love’, or emoticons.

7. Vowels can be removed and so must be inferred from context.

8. Punctuation, other than exclamation marks, is quite rare.

Read over your texts, forum messages and tweets and answer the following questions about each line.

1. How is it different to standard written English?

2. Which features from the above list can you identify?

3. What slang has been used?

4. Where have words been left out?

5. Would the meaning be clear to anyone or does the reader need to know something about the context? What would need to change to make it really clear?

Speech

Look carefully at your transcripts. Consider each line. How is the language used different to what you would expect if it had been a written text instead? Think about the following features of speech: repairs, hesitations, pauses, fillers, hedges, digressions, deletion, ellipsis, body language, phatic communication, repetition, replacement of words, slang, nicknames, abbreviations, and courtesy items. Which of these can you identify? Make notes on your transcript using your word processor’s ‘comments’ feature.

Comparison

Now that you have looked at text/online language and spoken language separately, it is time to compare them. Find as many similarities and differences as you can. What are the features of both types of language? How are they different to written language? What does a listener/reader need to know to understand each type? Can you see one of them changing very quickly, for example, as predictive text improves? What part does punctuation play?