cited from: http://www.geocities.com/davidvalentine2002/resources/tapelog.html
i. Notes on Transcription
There are many ways to transcribe a tape of an interview, or of naturally occurring conversation. The method you choose -- and the data you want to record -- depends on the kinds of data you want to elicit. For example, if you are a linguist or linguistic anthropologist, you may want to record the length of vowels, pitch, volume, or other para linguistic data. This could be very important if you are trying to investigate, for example, claims about "women's speech" which frequently fix on such markers. If you are more interested in the content of the interview/conversation, these features will be of less importance to you, but other features may be more important, such as noting the use of particular kinds of terms.
There are also different approaches to the amount of data one transcribes. In some cases, you may wish to transcribe the entire interview/conversation. In others, you may only want to transcribe the pieces that are the most relevant to you. Bear in mind, however, that the process of qualitative research is one of ongoing clarification of ideas and theses – i.e. those things that may not seem important to you right now may become of great importance later on.
My suggestion to most students is to transcribe tapes using a tape log. This affords you the most flexibility and gives you access to both context, linguistic, and para-linguistic (i.e. stuff that doesn't show up in transcription, like volume or pitch, but still important) data.
Above all, you should remember that your transcriptions are only ever a guide to your tape recorded data – if you end up using data from tapes in your work, you should go back and carefully check the transcriptions with the tapes. Hence, it is vital to record where on a tape the transcriptions come from (see below).
ii. Doing a tape log
A. Data to record up front.
To start your tape log, the following are vital to note on your transcriptions (on all pages – use a header in your transcription log to carry this information across pages):
1. The number of the tape (each of your tapes should be numbered, with the date of the event, the names of the particpants etc.; I also recommend keeping a log of the tapes you have)
2. The name(s) of interviewee(s)/conversational participant(s) (and pseudonym(s) if you are using them; alternatively, if you are very concerned about the identity of your interviewee, or participants in the conversation, record only their pseudonym and refer it to a list of pseudonyms that you keep separately in a password protected file; you can also password-protect your tape logs if you are concerned about privacy.)
3. Date and location of interview/conversation
4. Type of event (e.g. interview, life history, conversation, support group meeting)
5. PRODUCT NAME AND MODEL # OF TRANSCRIPTION DEVICE. (The latter two points are vital, because there is no standard for tape counters, and when you want to go back to find the place on the tape where the transcription comes from, you will need to have the same machine – this will save you tons of time).
6. Any notes that are pertinent to the event or your transcription (see example).
B. Creating a Tape Log
The most useful method I have found is to create a table in my word processing program with three columns. I title them: LOG, THEME/CONTEXT, and QUOTES.
The first column is to record the location of the transcribed information on the tape, and can be quite small. The second is to record the theme and context of the information. I use this to summarize the general themes, my impressions, important notes, para linguistic cues, contextual information etc.
The third column contains transcribed speech, as accurately as possible (more on this below).
C. Transcription conventions
Again, the kinds of conventions you use depend on the data you want. Some linguists find it of great importance to time pauses or silences in speech. Others focus on overlap between speakers, especially if they are interested in turn taking conventions. Others record pitch and volume. Whatever you choose to transcribe, remember that transcription is a time-consuming process that requires a good deal of attention and concentration.
For the purposes of most general ethnography, unless you are a linguistic anthropologist, I would strongly suggest you at least pay attention to the following, and use the following conventions to transcribe:
1. Note the speaker, first using their name/pseudonym, and thereafter, initials. Record your own speech and interjections, no matter how silly they sound. Note yourself by your own initials or by using "ME" or some such thing. If you're not sure who the speaker is, indicate this in square brackets with a question mark.
2. Use [square brackets] to note indistinct speech, or words that you think you can hear, but are not absolutely sure about. If you can't make head nor tail of the speech, leave empty brackets, indicating by the length of space the approximate space taken up by the indistinct speech.
3. Indicate overlap between speech by spacing. It is conventional to use square brackets between lines to show overlap e.g.
DV Tell us about yourself
[
Ben: Let me tell you about myself.
The square brackets indicate the point at which the overlapping speech starts. Sometimes, especially with multiple speakers, this can be an arduous task, but it is frequently useful to do.
4. Indicate pauses and silences – SILENCE AND PAUSES ARE DATA TOO! Likewise, transcribe hesitations such as "uh," "um," "er," etc. You may very well find, in looking back on your transcriptions, that pauses, hesitations, or silences are as telling as the speech itself -- why has someone stopped speaking? What comes after the silence? What might be indicated by the silence? Why has someone hesitated before using a certain word or voicing a certain opinion? Some linguists record the exact time of pauses (to the tenth of a second!) – this is usually not that important for most ethnographic purposes, but use ellipses for short pauses, and give information on long pauses with approximate time in {curly brackets} or [square brackets].
5. If paralinguistic data is useful, indicate it in square or curly brackets e.g. [everyone laughs], [very rapid speech]. Notable volume increase or shouting can be indicated by using ALL CAPS.
Exclamations can be marked with exclamation points! and emphasis with italics.
6. Accent and "eye-dialect." One of the most difficult things to note is accent and dialect. This is particularly the case with non-standard English accents. There has been an evolving critique of the representation of non-standard accents in linguistics. Basically, it boils down to this: those whose speech is represented in writing as non-standard are generally those who occupy lower socioeconomic positions, or have racially marked identities.
One example might be the representation of the word "can't" spoken by a white British English speaker and a black Jamaican working class English speaker. While the white middle class British English speaker doesn't produce "can't" with the American standard English pronunciation (a flat "a"), her or his speech is much less likely to be noted as "deviant" by a (white, middle class) American transcriber than the Jamaican speaker. So, a white middle class British English speaker may be noted as saying, "I can't do it, man," while the Jamaican speaker is more likely to be recorded as "I cay-an't do it, mon." (This latter transcription technique has been called "eye-dialect," meaning that it attempts to reproduce non-standard accents through orthography). If the same standards were held to the British English speaker, her or his words would be recorded as "I cahn't do it, man." This points to the fact that certain kinds of accents are marked – and this marking is linked to judgements about racial identification and socioeconomic status.
The solution is not easy, and there is not much consensus on how to record non-standard varieties. In many cases, some non-standard forms are now widely shared and thus are not as politically loaded (e.g., the dropping of the final "g" as in "happenin'"; the use of words such as "ain't" or "gonna"). In other cases, non-standard uses, for example, in African American Vernacular English, are widespread and accepted uses and should be noted. The most obvious e.g. of this is "aks" for "ask." Likewise, you should not necessarily "correct" word order, grammar, or mispronunciations in your transcripts. But neither should you strive to represent non-standard accents with non-standard spellings.
The point is that you should look to your own assumptions when you transcribe non-standard speech. You may want to note explicitly that a person has a particular kind of accent – but in doing so, you should be aware of the politics of such a noting, highlighting the fact that some of your participants speak with standard accents/dialects. In short, you should avoid attempting to represent accents, and use standard English spellings in transcriptions. The overall question is: why is there a standard in the first place; whose interests does this serve; and why do some people get represented as non-standard (e.g. African American speakers) while others don't (e.g. British English speakers, speakers of the "Boston Brahmin" dialect)?
For a good discussion of this, see:
Preston, Dennis R.
1982 'Ritin 'Fowklower Daun 'Rong: Folklorists' failure in phonology. Journal of American Folklore 95(377):304-326.
D. Recording Theme/Context
Use this column to introduce the theme of the section, the questions you've asked, stuff going on around you, interactions that aren't evident from the transcriptions, new participants arriving or present ones leaving, your own thoughts, notes etc. This is also a useful place to sum up the general theme of the speech or conversation. Indicate a change of theme by creating a new table row, and give each one a title. I generally use ALL CAPS to indicate a general theme.
E. Logging the Location of Speech on the Tape
This is an important part of the process. Indicate the location of the beginning of the thematic section in the LOG column, and preferably, the entire range of the thematic section. Further, it is also very useful to record the tape range next to transcribed speech. Also, in a long thematic section, you may want to note in the LOG column where particular new sub-themes are introduced. Although this may seem pointless to you right now, you will find when you have to go back to your tapes that you are very grateful to yourself for having done this!
F. Idiosyncracies
There are plenty of other conventions for transcription, but these are the ones that are most likely of use to most ethnographers. You will most likely find that you develop your own style as you transcribe (e.g. I choose to embolden the text I think will be most important for me, or the issues I find most important). If there are other issues you want to note, choose a convention. Most important of all, note the conventions that you have chosen to transcribe your tapes. This is important because you will not remember what they are a year or two later! A good place to do this is at the top of the tape log, where they will be the most likely to be seen.
iii. Conclusion
Again, transcription is an arduous process, but it is also very important. In representing people's voices in text, you are taking on a complex task, but as an ethnographer, this is perhaps the strongest data you will have to make your arguments and to give voice to your study participants.
One final -- and vital -- thing to remember: for all the work you're going to put into transcribing your tapes, transcriptions are still only ever representations. Even if you accurately record every single pause, every intonation, pitch, and volume change, every overlap and interjection, what you have produced is still a representation – it is not the "truth." You can never completely and accurately record the gestalt of the event. But it is an important step in attempting to represent, as accurately as possible, the speech (and by extension, the self-representations) of the people with whom you have been privileged to work as an ethnographer.
i. Notes on Transcription
There are many ways to transcribe a tape of an interview, or of naturally occurring conversation. The method you choose -- and the data you want to record -- depends on the kinds of data you want to elicit. For example, if you are a linguist or linguistic anthropologist, you may want to record the length of vowels, pitch, volume, or other para linguistic data. This could be very important if you are trying to investigate, for example, claims about "women's speech" which frequently fix on such markers. If you are more interested in the content of the interview/conversation, these features will be of less importance to you, but other features may be more important, such as noting the use of particular kinds of terms.
There are also different approaches to the amount of data one transcribes. In some cases, you may wish to transcribe the entire interview/conversation. In others, you may only want to transcribe the pieces that are the most relevant to you. Bear in mind, however, that the process of qualitative research is one of ongoing clarification of ideas and theses – i.e. those things that may not seem important to you right now may become of great importance later on.
My suggestion to most students is to transcribe tapes using a tape log. This affords you the most flexibility and gives you access to both context, linguistic, and para-linguistic (i.e. stuff that doesn't show up in transcription, like volume or pitch, but still important) data.
Above all, you should remember that your transcriptions are only ever a guide to your tape recorded data – if you end up using data from tapes in your work, you should go back and carefully check the transcriptions with the tapes. Hence, it is vital to record where on a tape the transcriptions come from (see below).
ii. Doing a tape log
A. Data to record up front.
To start your tape log, the following are vital to note on your transcriptions (on all pages – use a header in your transcription log to carry this information across pages):
1. The number of the tape (each of your tapes should be numbered, with the date of the event, the names of the particpants etc.; I also recommend keeping a log of the tapes you have)
2. The name(s) of interviewee(s)/conversational participant(s) (and pseudonym(s) if you are using them; alternatively, if you are very concerned about the identity of your interviewee, or participants in the conversation, record only their pseudonym and refer it to a list of pseudonyms that you keep separately in a password protected file; you can also password-protect your tape logs if you are concerned about privacy.)
3. Date and location of interview/conversation
4. Type of event (e.g. interview, life history, conversation, support group meeting)
5. PRODUCT NAME AND MODEL # OF TRANSCRIPTION DEVICE. (The latter two points are vital, because there is no standard for tape counters, and when you want to go back to find the place on the tape where the transcription comes from, you will need to have the same machine – this will save you tons of time).
6. Any notes that are pertinent to the event or your transcription (see example).
B. Creating a Tape Log
The most useful method I have found is to create a table in my word processing program with three columns. I title them: LOG, THEME/CONTEXT, and QUOTES.
The first column is to record the location of the transcribed information on the tape, and can be quite small. The second is to record the theme and context of the information. I use this to summarize the general themes, my impressions, important notes, para linguistic cues, contextual information etc.
The third column contains transcribed speech, as accurately as possible (more on this below).
C. Transcription conventions
Again, the kinds of conventions you use depend on the data you want. Some linguists find it of great importance to time pauses or silences in speech. Others focus on overlap between speakers, especially if they are interested in turn taking conventions. Others record pitch and volume. Whatever you choose to transcribe, remember that transcription is a time-consuming process that requires a good deal of attention and concentration.
For the purposes of most general ethnography, unless you are a linguistic anthropologist, I would strongly suggest you at least pay attention to the following, and use the following conventions to transcribe:
1. Note the speaker, first using their name/pseudonym, and thereafter, initials. Record your own speech and interjections, no matter how silly they sound. Note yourself by your own initials or by using "ME" or some such thing. If you're not sure who the speaker is, indicate this in square brackets with a question mark.
2. Use [square brackets] to note indistinct speech, or words that you think you can hear, but are not absolutely sure about. If you can't make head nor tail of the speech, leave empty brackets, indicating by the length of space the approximate space taken up by the indistinct speech.
3. Indicate overlap between speech by spacing. It is conventional to use square brackets between lines to show overlap e.g.
DV Tell us about yourself
[
Ben: Let me tell you about myself.
The square brackets indicate the point at which the overlapping speech starts. Sometimes, especially with multiple speakers, this can be an arduous task, but it is frequently useful to do.
4. Indicate pauses and silences – SILENCE AND PAUSES ARE DATA TOO! Likewise, transcribe hesitations such as "uh," "um," "er," etc. You may very well find, in looking back on your transcriptions, that pauses, hesitations, or silences are as telling as the speech itself -- why has someone stopped speaking? What comes after the silence? What might be indicated by the silence? Why has someone hesitated before using a certain word or voicing a certain opinion? Some linguists record the exact time of pauses (to the tenth of a second!) – this is usually not that important for most ethnographic purposes, but use ellipses for short pauses, and give information on long pauses with approximate time in {curly brackets} or [square brackets].
5. If paralinguistic data is useful, indicate it in square or curly brackets e.g. [everyone laughs], [very rapid speech]. Notable volume increase or shouting can be indicated by using ALL CAPS.
Exclamations can be marked with exclamation points! and emphasis with italics.
6. Accent and "eye-dialect." One of the most difficult things to note is accent and dialect. This is particularly the case with non-standard English accents. There has been an evolving critique of the representation of non-standard accents in linguistics. Basically, it boils down to this: those whose speech is represented in writing as non-standard are generally those who occupy lower socioeconomic positions, or have racially marked identities.
One example might be the representation of the word "can't" spoken by a white British English speaker and a black Jamaican working class English speaker. While the white middle class British English speaker doesn't produce "can't" with the American standard English pronunciation (a flat "a"), her or his speech is much less likely to be noted as "deviant" by a (white, middle class) American transcriber than the Jamaican speaker. So, a white middle class British English speaker may be noted as saying, "I can't do it, man," while the Jamaican speaker is more likely to be recorded as "I cay-an't do it, mon." (This latter transcription technique has been called "eye-dialect," meaning that it attempts to reproduce non-standard accents through orthography). If the same standards were held to the British English speaker, her or his words would be recorded as "I cahn't do it, man." This points to the fact that certain kinds of accents are marked – and this marking is linked to judgements about racial identification and socioeconomic status.
The solution is not easy, and there is not much consensus on how to record non-standard varieties. In many cases, some non-standard forms are now widely shared and thus are not as politically loaded (e.g., the dropping of the final "g" as in "happenin'"; the use of words such as "ain't" or "gonna"). In other cases, non-standard uses, for example, in African American Vernacular English, are widespread and accepted uses and should be noted. The most obvious e.g. of this is "aks" for "ask." Likewise, you should not necessarily "correct" word order, grammar, or mispronunciations in your transcripts. But neither should you strive to represent non-standard accents with non-standard spellings.
The point is that you should look to your own assumptions when you transcribe non-standard speech. You may want to note explicitly that a person has a particular kind of accent – but in doing so, you should be aware of the politics of such a noting, highlighting the fact that some of your participants speak with standard accents/dialects. In short, you should avoid attempting to represent accents, and use standard English spellings in transcriptions. The overall question is: why is there a standard in the first place; whose interests does this serve; and why do some people get represented as non-standard (e.g. African American speakers) while others don't (e.g. British English speakers, speakers of the "Boston Brahmin" dialect)?
For a good discussion of this, see:
Preston, Dennis R.
1982 'Ritin 'Fowklower Daun 'Rong: Folklorists' failure in phonology. Journal of American Folklore 95(377):304-326.
D. Recording Theme/Context
Use this column to introduce the theme of the section, the questions you've asked, stuff going on around you, interactions that aren't evident from the transcriptions, new participants arriving or present ones leaving, your own thoughts, notes etc. This is also a useful place to sum up the general theme of the speech or conversation. Indicate a change of theme by creating a new table row, and give each one a title. I generally use ALL CAPS to indicate a general theme.
E. Logging the Location of Speech on the Tape
This is an important part of the process. Indicate the location of the beginning of the thematic section in the LOG column, and preferably, the entire range of the thematic section. Further, it is also very useful to record the tape range next to transcribed speech. Also, in a long thematic section, you may want to note in the LOG column where particular new sub-themes are introduced. Although this may seem pointless to you right now, you will find when you have to go back to your tapes that you are very grateful to yourself for having done this!
F. Idiosyncracies
There are plenty of other conventions for transcription, but these are the ones that are most likely of use to most ethnographers. You will most likely find that you develop your own style as you transcribe (e.g. I choose to embolden the text I think will be most important for me, or the issues I find most important). If there are other issues you want to note, choose a convention. Most important of all, note the conventions that you have chosen to transcribe your tapes. This is important because you will not remember what they are a year or two later! A good place to do this is at the top of the tape log, where they will be the most likely to be seen.
iii. Conclusion
Again, transcription is an arduous process, but it is also very important. In representing people's voices in text, you are taking on a complex task, but as an ethnographer, this is perhaps the strongest data you will have to make your arguments and to give voice to your study participants.
One final -- and vital -- thing to remember: for all the work you're going to put into transcribing your tapes, transcriptions are still only ever representations. Even if you accurately record every single pause, every intonation, pitch, and volume change, every overlap and interjection, what you have produced is still a representation – it is not the "truth." You can never completely and accurately record the gestalt of the event. But it is an important step in attempting to represent, as accurately as possible, the speech (and by extension, the self-representations) of the people with whom you have been privileged to work as an ethnographer.
No comments:
Post a Comment