Transcription rules
This article is an excerpt from our eBook Recording, Typing, Analysing - Guide to Conducting Interview & Transcription.
The book is available as Free downloadFind out everything about transcription & co now!
Table of contents of this article
- What are transcription rules?
- Simple and Scientific Transcription
- Advanced transcription
- Complex transcription
What are transcription rules?
There are many different transcription rules or transcription systems. The Transcription rules depend on the purpose for which the transcript is made.transcriptions are made for the public (e.g. newspaper articles, films, series) or for studies (market and opinion research, sociological or linguistic) and what is to be examined (the content or the language?) and how detailed the evaluation should be.
In principle, a distinction is usually made between Simple and extended transcription rules differentiation.
The simple transcription is recommended above all when the content is to be analysed. The transcription is word-for-word accurate but the language is slightly smoothed. This means that stuttering, word breaks and hesitation sounds ("äh", "ähm") are omitted from the transcription. This increases readability and makes the content more accessible. The simple transcription rules can also vary depending on the author.
From years of experience we have therefore abtipper.de's own transcription rules transcription rules. These are presented in detail on this page.
The transcription rules mainly concern precise instructions on how to notate what is said. In a separate article, you will find detailed information on the recommended Formatting a transcript.
The transcription rules of abtipper.de are scientific quotable as:
Claussen, J. / Jankowski, D. / Dawid, F.: Aufnehmen, Abtippen, Analysieren - Wegweiser zur Durchführung von Interview und Transkription; Hannover 2020, ISBN: 978-3750470057
The transcription rules of abtipper.de se are similar to the Transcription rules by Dresing and Pehl. Both the simple transcription rules and the extended transcription rules of abtipper.de are a Literal transcription.
In contrast to simple transcription systems, the language is not Advanced transcription is not smoothed. This means that word breaks, stutterers, hesitation sounds ("äh", "ähm") and intermediate sounds (e.g.: "hm") are also transcribed. Extended procedures are particularly suitable when linguistic aspects are also examined. However, readability decreases the more detailed the transcript is prepared.
Very detailed transcripts are produced according to complex transcription rules for studies in the fields of social sciences and linguistics. In the field of social sciences, the Transcription methods according to Bohnsack, TIQ is often used.
Frequently asked are the Mayring's transcription rules. Mayring is particularly known for his guides to content analysis. The "Note on Interview Transcription" from Mayring (2015, p. 57) describes fragments of applied transcription rules of a specific project presented in this work and is not an official, scientific transcription procedure. If you nevertheless wish to transcribe according to these rules, we offer a transcription procedure following this notation system. Since most aspects correspond to those of our rules, only the coding of pauses and incomprehensible passages must be modified.
In contrast to the simple transcription rules of abtipper.de, which provide for the setting of a time stamp at an incomprehensible passage, Mayring indicates such a passage by bracketed dots. The number of these dots represents the approximate length of the incomprehensible passage. Pauses of 4 seconds or more are indicated by a number of seconds in brackets according to the simple rules of abtipper.de, whereas Mayring's rules provide for the notation of dashes in this case. The number of dashes again represents the length of the pause.
If you would like to order a transcription based on Mayring, select "scientific transcription" when ordering and write a short note about Mayring in the field for special requests.
In linguistics, the transcription rules are even more complex. Among the best known are the phonetic transcription (according to the IPA), HIAT and GAT2 (conversation analytical transcription system).. For GAT2, a basic transcript, fine transcript and minimal transcript can be made. Which and how much should be taken into account again depends on the research question. Transcription according to complex transcription rules is usually very time-consuming; the more aspects are to be taken into account, the more time should be planned. On request, we can prepare your minimal transcript as well as transcripts according to HIAT.
In general, transcription rules can be extended by many options. The options that can be selected include time stamps, line numbering or anonymisation. Timestamps and line numbering help to find certain passages in the text more easily. Time stamps can be inserted as needed, e.g. every minute or by default after each change of speech. Anonymisation can also be done individually. For example, names, institutions or cities can be anonymised.
Another option is to make a smoothed transcription. This is recommended, for example, for the field of media and journalism. The result is a ready-to-print text that is particularly easy to read. To achieve this, some rephrasing is done, e.g. in the case of repetitive sentence beginnings, and word and sentence breaks are omitted. Smoothing is therefore less suitable for a scientific paper.
There are a variety of transcription procedures and rules, mostly from the fields of social and linguistic sciences. The individual procedures differ above all in terms of their complexity (simple, extended, complex) and their range of application. Almost every procedure can be refined or extended according to one's own needs. Especially within the complex procedures like GAT2, some parameters are optional (e.g.: parameters regarding volume or speaking rate).
Transcription procedures differ in terms of complexity - individual aspects are optional.
Simple and Scientific Transcription Rules
For many purposes, transcription according to simple rules is the most suitable procedure. This is especially true when the content is the main focus of the analysis, e.g. interviews for the public, such as the press or film and television, but also for scientific questions outside linguistics, such as economics or market research. A transcription made using a simple procedure is also easy to smoothedthat is, stutterers, slips of the tongue and intermediate sounds such as "äh" and "ähm" are not taken into account. Dialectal utterances are also rendered in standard language. This means that the transcript is easy to read and can be passed on to the public so that, for example, interviews can be published on online sites or in print media.
Simple procedures are recommended when content-related aspects are in the foreground - the transcript is smoothed slightly and is thus particularly accessible, also for the public
A scientific transcription is edited according to the rules of simple transcription and additionally checked by a Editor proofreader. This is particularly suitable for theses.
In the case of scientific transcription, the transcript is additionally checked by a proofreader
Based on years of experience, we at abtipper have developed our own transcription rules, following simple and advanced procedures.
Simple transcription rules from abtipper.de
These following simple transcription rules from abtipper.de are scientifically citable as:
Claussen, J. / Jankowski, D. / Dawid, F.: Aufnehmen, Abtippen, Analysieren - Wegweiser zur Durchführung von Interview und Transkription; Hannover 2020, ISBN: 978-3750470057
The following guidelines should be followed for the simple transcription:
- The text is adopted as it is spoken. No corrections are made, i.e. errors (e.g. grammatical errors in sentence order) are adopted. Exceptions: See points 3 to 5.
- All statements, even seemingly unimportant filler words (e.g. "I say times" or "so to speak" etc.) are recorded.
- Dialect colourings are corrected (e.g. "haben wir" instead of "hamma").
- All non-verbal intermediate sounds of the speakers (e.g. stutterers, uhms, ne?) are omitted.
- All listener acknowledgements irrelevant to the content (e.g. Hm, Ja, Ach ja) are also omitted. These are only transcribed in rare cases when these words contribute to the content (e.g. in response to a question).
- Special events are placed in brackets (e.g. (sound disturbance) or (telephone rings several times)).
- Abbreviations are only used if the person pronounces them in the same way (e.g. a spoken "et cetera" is not abbreviated with "etc." in the transcript).
- Only literal/direct speech is put in inverted commas (e.g. I asked him: "Why are you doing that?").
- To avoid tapeworm sentences over several lines, punctuation marks are placed sensibly. A conjunction (e.g. "and") can be placed at the beginning of a sentence.
- Polite pronouns such as "you" and "yours" are capitalised. For example, when people address each other during an interview (e.g. "I have another question for you."). If the interview partners are on a first-name basis, the "du" and all forms of the "du" (i.e. also: "dir", "dich", "dein", etc.) are written in lower case.
- The capitalisation of foreign words is chosen as one would write the German equivalent, i.e. verbs lower case and nouns upper case (e.g. I googled cyberspace.).
- All numbers from one to twelve are written out in full and 13 and above are written as digits. Meaningful exceptions such as the date are also written as digits (i.e. "3.1.2017").
- Especially important for an exact and quick assignment: The transcript is given the exact file name of the audio file (e.g. "REC- 0005"). If only one section was transcribed, the corresponding minutes are added to the file name (e.g. "REC-0005 - Minute 0-30").
- The interviewer is named I and the interviewee is named B. If there is more than one person, a number is added, e.g. I1, I2, B1 etc. The names of the persons are written in bold. Exceptions to this are insertions (see point 21).
- Incomplete sentences are marked with a "-" (e.g. "So then there were-, no, again: there were four people in the-."). Regular punctuation marks are placed after the "-" as in the example. The "-" is placed directly after the word, without a space.
- Incomplete words are only included if they have added content value. Otherwise, they are considered stutterers and are simply omitted.
- Pauses longer than four seconds are marked with the number of seconds in brackets, e.g. for a pause of seven seconds: (7 sec.).
- Words where the wording is not quite clear and is only assumed are marked with a question mark and put in brackets (e.g. (?luminary)). If the understood word obviously makes no sense and cannot logically fit in at all at this point, then the passage is marked as incomprehensible (see next point).
- Unintelligible passages (e.g. due to noise or other interference) are marked with a time stamp according to the format ... #hh:mm:ss#. In the case of ...#00:01:04# there would therefore be an incomprehensible passage after 1 minute 4 seconds.
- With the exception of item 19, no time stamps are set in the simple transcription.
- In the case of very short insertions by the other person (even spoken at the same time), these statements can be incorporated into the other person's flow of speech in parentheses (e.g. "I: That was 12 years, (B: No, 13.) I remember."). This does not apply to listener acknowledgements without added content value, which are simply omitted (e.g. Hm). In the case of insertions, the speaker designations are not written in bold. Insertions are also ended with a punctuation mark, usually a full stop. Other punctuation marks are placed before the insertion, not after it.
Example of a transcript according to simple transcription rules:
Name of the file: Interview Mr Müller v2
I1: Yes, how was that for you?
B: Well, from the body sensation, the other experience had been more intense.
I1: In what way?
B: -I can't judge 100 per cent, because I wasn't completely 100 per cent there for the whole eight minutes. So, I say, it could also have developed differently, (I1: I see it differently.) if I-. (4 sec.)
I1: Kind of. (Phone rings.)
B: Yes, well, it is somehow very, very strange because I normally never fall asleep quickly. I have been to many therapies and so on, etc., for example last year in Cologne with Mr (?Schindlorz). He also asked me: "How can that be?"
I2: You said just now, when you came, that you were very ... #00:01:47#. Could you explain that again?
Advanced transcription rules
Extended procedures are particularly suitable if a detailed evaluation is to be carried out in which, in addition to content-related aspects, the following are also to be taken into account Linguistic aspects are to be taken into account. These include special verbal aspects, such as stutterers, as well as prosodic aspects (special stresses). Thus, transcription according to the extended method is more time-consuming than according to simple rules. At the same time, the readability of the transcript is made more difficult for outsiders, so that the extended procedure is only better suited in a few cases.
In extended procedures, linguistic aspects are also taken into account, which makes transcription more time-consuming - extended procedures are only recommended in rare cases.
Advanced transcription rules from abtipper.de
The following extended transcription rules from abtipper.de are scientifically citable as
Claussen, J. / Jankowski, D. / Dawid, F.: Aufnehmen, Abtippen, Analysieren - Wegweiser zur Durchführung von Interview und Transkription; Hannover 2020, ISBN: 978-3750470057
The following guidelines should be followed for the extended transcription:
- The text is adopted as it is spoken. No corrections are made, i.e. errors (e.g. grammatical errors in sentence order) are adopted.
- All statements, even seemingly unimportant filler words (e.g. "I say times" or "so to speak" etc.) and intermediate sounds of the speakers (e.g. stutterers, Hms, Ähms etc.), are adopted.
- Dialect colourings are corrected (e.g. "haben wir" instead of "hamma").
- Special events are placed in brackets (e.g. (sound disturbance) or (telephone rings several times)).
- Particularly stressed terms are capitalised (e.g. "We will NOT do that.").
- Abbreviations are only used if the person pronounces them in the same way (e.g. a spoken "for example" is not abbreviated to "e.g." in the transcript).
- Verbatim/direct speech is regularly put in inverted commas (e.g. I asked him: "Why are you doing that?").
- To avoid tapeworm sentences over several lines, punctuation marks are set sensibly.
- Polite pronouns such as "you" and "yours" are capitalised. For example, when people address each other during an interview (e.g. "I have another question for you."). If the interview partners are on a first-name basis, the "du" and all forms of the "du" (i.e. also: "dir", "dich", "dein") are written in lower case.
- All numbers from one to twelve are written out and all numbers from 13 onwards are written as digits.
- The interviewer is named I and the interviewee is named B. If there is more than one speaker, a number is added, e.g. I1, I2 etc.
- The name of the persons is written in bold.
- Incomplete sentences are marked with a "-" (e.g. "So then there were-, no, let's start again: There were four people.").
- Words where the wording is not quite clear and is only conjectural are marked with a question mark and placed in brackets (e.g. (?luminary)).
- Pauses longer than four seconds are placed in brackets with the number of seconds, e.g. for a pause of seven seconds: (7 sec.).
- Unintelligible passages (e.g. due to noise or other interference) are marked with a time stamp according to the format ... #hh:mm:ss#. In the case of ...#00:01:04# there would therefore be an incomprehensible passage after 1 minute 4 seconds.
- After each speaker change, a timestamp in the format #hh:mm:ss# is inserted.
- In the case of very short interjections by the other person (also listener acknowledgements and what is spoken at the same time), e.g. in an interview, this statement is included in the other person's flow of speech in brackets (e.g. "I: I was new here at the time (B: I see.) and therefore didn't know many."). In these insertions, the speaker's names are not written in bold.
Example of a transcript according to extended transcription rules:
Name of the file: Interview Mr. Müller v2
I1: Yes, um, so how was that for you? #00:00:01#
B: So, so, so from the body sensation the other experience had been more intense. #00:00:03#
I1: In what way? #00:00:10#
B: -not so 100 percent um judge, because I wasn't completely 100 percent there for the whole eight minutes. So, I say, it could also have developed differently (I1: Exactly.), if I had been awake-. (4 sec) So on the other hand, it's also a good sign that I was completely relaxed. #00:01:07#
I1: Kind of. (Phone rings.) #00:01:16#
B: Hm (asking). So it's kind of VERY VERY strange because I usually never fall asleep quickly. I have been in many therapies and so on, for example last year in Cologne with Mr (?Schindlorz). They also asked me: "How can that be?" #00:01:45#
I2: You just said that when you came, you were very ... #00:01:47#. Could you elaborate on that again? #00:01:50#
Complex transcription rules
The complex procedures include TiQ, HIAT and GAT2, among others. These procedures are so complex that they are usually only used in the social sciences and linguistics. Thus, they are only used in very specific fields of application.
Complex procedures can have different emphases. These are not only, as with simple and extended transcription, on content and verbal aspects. Special attention is paid to the exact reproduction of what is said. Non-verbal and prosodic features are also taken into account. The complex procedures are suitable for gaining an understanding of the transcript when reading it. Auditory impression to gain a better understanding of the transcript. The more parameters are set, the more can be analysed and also interpreted; accordingly, the analysis is more elaborate and takes more time with complex procedures.
Complex procedures are only used in very specific fields of application - among other things, they serve to convey an auditory impression.
A well-known complex procedure is the TIQ procedure. The TiQ method (according to Bohnsack) is primarily geared towards sociological research questions. Compared to HIAT and GAT2, the TiQ method is more accessible (the reason for this is also the presentation). However, TiQ is not suitable for linguistic research.
TiQ transcription rules
- Literal transcription; intermediate sounds, listener acknowledgements ("uh", "huh" etc.) and emotional expressions ("laugh") are taken over
- Words are capitalised at the beginning of the utterance and at the beginning of an overlap, after a └ . After punctuation marks, however, words are written in lower case, since punctuation marks are intonational and not grammatical in nature. An exception is also made for nouns, which are also capitalised.
- Lines are numbered
- All participants are assigned a letter with the suffix f for females and m for males (e.g.: Af, Bm, Cf).
Other signs and symbols in the TiQ procedure:
- └ Start of an overlap
- ┘End of an overlap
- (.) Pause up to one second
- (2) Number of seconds of a speaker pause
- Emphasised is underlined
- What is saidaloud is written in bold
- °quiet° speaking is marked with °.
- . sharply falling intonation
- ; Slightly sinking intonation
- ? strongly rising intonation
- , weakly rising intonation
- - indicates termination of a word: lei-
- = marks word blends: ham=ma
- : marks elongation of vowels, the frequency corresponds to the length of the elongation e.g.: "nei::n"
If there is uncertainty about the exact wording, the word is put in brackets, e.g.: (but)
() incomprehensible utterances, the length of the bracket corresponds approximately to the duration of the incomprehensible utterance
((groans)) Comments or remarks on para-linguistic, non-verbal or conversation-external events; the length of the bracket in the case of commenting on para-linguistic utterances (e.g. moaning) corresponds approximately to the duration of the utterance.
@no@ e.g. laughing spoken "no"
@(.)@ short laugh
@(3)@ 3 seconds of laughter
//mhm// Listener signal from the interviewer if the "mhm" is not overlapping.
HIAT and GAT2 are complex, individually extendable procedures that are mainly used in the field of linguistics. With these methods it can even be useful to work with Video material because HIAT and GAT2 also take into account non-verbal communication and action.
It can also be useful to work with video material in complex procedures - in this way non-verbal communication can be analysed.
The HIAT method has some advantages, especially when several speakers communicate at the same time and when further prosodic features are to be marked. The Score notation does affect readability, but it allows several aspects to be illustrated vividly and clearly.
For complex procedures such as HIAT, a score notation is often specified - a corresponding programme such as EXMARaLDA is recommended for the representation
If the transcription is to be made according to the HIAT procedure, it is therefore advisable to work with EXMARaLDA. EXMARaLDA is a linguistic system with tools for creating and analysing corpora of conversations. This includes the Score Editor tool for making transcripts. Below is an example of the representation as a score in EXMARaLDA:
HIAT and GAT2 give a better auditory impression, but the transcripts become increasingly illegible with increasing volume. Furthermore, these methods take more time because each conversation excerpt has to be checked several times for different phenomena (such as pauses, main accents, pitch progressions, etc.).
GAT2, originally GAT (Gesprächsanalytisches Transkriptionssystem), was developed by linguists with the aim of creating a uniform system. This was to enable data from different research directions to be analysed. The revised version GAT2 has existed since 2009.
GAT2 is also mainly used in linguistics - it differentiates between minimal, basic and fine transcript
In GAT2, a distinction is made between three transcripts, which can be combined as desired: the Minimal transcript, the Basic transcript and the Fine transcript.
The minimal transcript contains information on the course structure, e.g. overlaps, simultaneous speaking and pauses.
In the basic transcript, turns are segmented into intonation phrases, the following aspects can be taken into account:
- Pitch movements at the end of the phrase (.,;-?)
- Focus accent and strong accent. Ex: ak!ZENT!
- Stretching sounds
- interpretierende Kommentare wie <<lachend>wow> etc.
In the fine transcript, secondary accents, accent pitch movements, pitch jumps, changes in volume/speed etc. are noted. The fine transcript is especially interesting for linguists in the field of conversation analysis/ intonation phonology.
For the GAT2 procedure, an equidistant font (such as Courier) should be used, as this is a condition for further processing of the transcripts (e.g. for simultaneous speech). Here is an example of a minimal transcript according to GAT2:
A detailed step-by-step guide for transcription according to GAT2 is available free of charge in PDF format from Hagemann/Henle (2014). There is also an online tutorial from the University of Freiburg with practical tips on transcribing according to GAT2:
http://paul.igl.uni-freiburg.de/gat-to/
| Transcription rules | Overview transcription rules | ||
| Simple rules from typing | Advanced Rules from typing | TIQ | |
| Word-for-word accuracy (errors are also taken over) | yes | yes | yes |
| Filler words ("so to speak", "I say") | yes | yes | yes |
| Reception signals (e.g., "mhm" affirmative). | yes, if answer to a question | yes | yes |
| Hesitation phenomena (e.g. "uh", "um") | no | yes | yes |
| Dialect | Standard German (exception: dialectal words without accurate translation) | Standard German (exception: dialectal words without accurate translation) | is adopted |
| Word and sentence breaks | Sentence breaks with - | both with - | Word break with - |
| Word slurring (e.g.: "I have it" instead of "I have it") | Standard German (I have it) | Standard German (I have it) | with = (I have=s) |
| Interjections (e.g.: "oh", "oops", "shush") | no | yes | yes |
| Speech pauses | from 4 seconds Number of seconds in brackets (5 sec) | from 4 seconds Number of seconds in brackets (5 sec) | Pause up to one second: (.); otherwise number of seconds in brackets (4) |
| Overlaps | Insertions (even if overlapping) in brackets: I: That was two years ago (B: No!), I remember | Insertions (even if overlapping) in brackets: I: That was two years ago (B: No!), I remember | () Length of the bracket corresponds to the duration of the unv. statement |
| short inserts | |||
| presumed wording | ( wording ) | ( wording ) | (Wording) |
| unintelligible wording | Marking at exact location with ... and a timestamp | Marking at exact location with ... and a timestamp | |
| verbatim speech | is placed in inverted commas | is placed in inverted commas | |
| Prosody | |||
| strong emphasis | no | in capitals | Underline |
| spoken loudly/ softly | no | no | louder/ °quieter° |
| Stretching a word | no | no | Stretching, frequency corresponds to length of stretching |
| Intonation | no | no | by punctuation marks (.;,?) |
| Non-verbal events (e.g.: disturbing noises, telephone ringing) | in brackets (phone rings) | in brackets (phone rings) | in double brackets ((phone rings)) |
| Para-linguistic events (e.g.: laughing, crying etc.) | in brackets (laughter) | in brackets (laughter) | in double brackets ((cry)), exception laugh with @ symbols: @(.)@ short laughing up |
| Upper and lower case | according to German grammar, polite pronouns ("Sie", "Ihre") start with a capital letter | according to German grammar, polite pronouns ("Sie", "Ihre") start with a capital letter | Everything is written in lower case except for nouns |
| Punctuation marks | according to the official German spelling rules, tapeworm sentences are avoided | according to the official German spelling rules, tapeworm sentences are avoided | s. Intonation |
| Timestamp | Only if the wording is incomprehensible | after each change of speaker and in case of unintelligible wording | n.a. |
Order your transcription now at abtipper.de!
Transcription rules determine the specifications according to which audio and video recordings are to be converted into text. You determine, for example, what must be noted and how, what can be omitted and what the finished text should look like. Transcript should look like.
A distinction is typically made between simple, extended and complex transcription rules. The latter two are used almost exclusively for scientific transcriptions.
Simple Transcription rules are the standard outside of academia. They specify a word-for-word transcription, but allow non-content-relevant non-verbal elements (e.g. stutterers) to be left out.
In the social sciences, simple transcription rules are often used for content analysis. In the linguistic sciences, on the other hand, extended or complex rules are common.
Advanced Transcription rules are mainly used in linguistics. They specify a word-for-word transcription including all non-verbal elements (e.g. stutterers).
In linguistics, there are also a number of complex transcription rules, such as GAT2, HIAT or TIQ, which in some cases make very elaborate and specific specifications for transcription and formatting.
Outside of linguistics, the use of these procedures is rather uncommon, as the resulting transcripts are often difficult to read and have only minor advantages in terms of content compared to the result of the simple transcription rules.
