On the next page I'll suggest specific guidelines on how to change names in transcripts, but first some general preamble.

If your data is from a broadcast source (for example, a television talk show, or a radio phone-in), then people have knowingly identified themselves to the world at large, and there would normally be no reason to anonymise them in a scholarly paper. (Indeed, it would be odd to do so, if they were politicians or other public figures).

On the other hand, if you have recorded the talk of people in private conversations, then you will have asked for their explicit informed consent. Increasingly, this is done with a consent form that people sign; and on that form, it is standard practice to include the promise that you will only publish their talk in a form that will not identify the individuals personally.

You will need to change people's names and any other identifying information - placenames, addresses and so on.

Why not just use labels like "Speaker 1", "Speaker 2" and so on? If it was just a matter of identifying who is speaking, you could change all names to simple labels (e.g. Teacher, Child 1, Child 2 and so on, or Speaker A, Speaker B and so on). However, that can impose categories on the speakers (which might be perfectly uderstandable in most cases, but bring with them unwanted or ungrounded implications in others, for example Friend 1, or Neighbour A). It also tends to make the transcript lifeless. The most compelling reasons against labels, though, are to do with how people are referred to in the talk. (a) labels used for names in the talk itself can lead to the reader losing track, especially when there are many participants, and speakers refer to one another in the third person. It's better to avoid such dialogue as "Miss, [Child 1] said [Child2] that he couldn't play with [Child 3] because [Child 2] didn't share his sweets with her". (b) more importantly, it is crucial in CA transcripts to be able to describe the delivery of talk exactly - so if there is an unusual intonation emphasis in the name Melissa, for example, or if a speaker overlapped the start of their turn with the middle of the previous speaker's reference to Doctor Gutierrez, then it's impossible to show that accurately if the label (Child A, Doctor 1) is of a different syllabic length or intonational pattern.

Pseudonyms add vividness But they must be used carefully. These are useful rough-and-ready guidelines to get you going. On some occasions, they will be trumped by other considerations, but for the most part, they will be serviceable.

Now you're ready to for the next page, which gives you ten guidelines on how to change names.