[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
If you are new to the field of sign linguistics, think carefully about what you want to investigate before you start collecting data. If not, you may find that you haven't collected the right material or that you didn't use the right camera positioning, etc.
Note that sign languages are often very different from spoken languages. Not only do they not have a written tradition or even an accepted writing system, such as western languages, but they are also acquired in a different way than children acquiring a spoken language (again, in western countries), since most deaf children do not have parents who use a sign language. As yet, there is usually no norm and little correction on children's sign language by parents or educators.
Also think carefully about the framework you are going to use. You should not take for granted that notions, frameworks and tools used in research on spoken languages are automatically fit for research on sign languages.
Before starting to elicit data, again consider carefully what you are looking for. In some cases it may be necessary to use written sentences that the informants should translate. In these case you should be aware of the influence of the spoken language on the data.
Often non-language based elicitation material is used, such as pictures, comics or video-clips. There is already much material, and for cross-linguistic comparison it is good to share these materials. For research into several fields (such as classifiers and use of space) there is already standardized material available.
The signing under investigation should, whenever possible, be recorded, preferably on digital video. Make sure that the signer is in full view of the camera (even with arms stretched out or high up). It is adviseable to use more than one camera (if possible). This is especially so for a close-up of the face, since a lot of information is present in/on the face. In case there is a second signer present (e.g. the interlocutor), it appears to be very handy to record him as well, in order to capture the interaction; sometimes (re)actions of the interlocutor or addressee have influence on the utterances of the signer.
Make sure that there is good light. The signer should preferrably not wear a red shirt, or a shirt with small stripes or checks because this may blur the view.
Before starting the real session it is clever to videotape a few minutes and check whether everything is OK. (You don't want to find out after your session that you left the lens cap on or something like that.)
The brightness of the recording will often look diffferently on the little screen of the video camera, on a TV monitor, and on a computer screen. It is therefore useful to make test recordings and go through the whole process (of digitisation, compression, etc.) to see what the image looks like in the set-up you will actually use for analysis later on. It is possible to change aspects like brightness and contrast later during compression, but this is not always easy, and it is best to avoid those kind of modifications.
In case you are not a fluent signer and don't know many signers, be careful in asking signers who would be a good informant. How could they know what you are looking for, if they are not trained in linguistic research? They may advise you someone who is a fluent signer in another sign language, or a signer who may be fluent, but who uses a form of speech-supported sign. In case there are no collegues to help you, the best thing to do in this case is to try to figure out the backgrounds of candidate informants, such as the age at which they became deaf, whether they have Deaf/signing family members and/or friends, what type of school they attended, what the language of instruction was at school and how they communicated with their (deaf) peers, and how often/regularly they use sign.
In case you have no clue you can film different signers. Learn/practice the sign language you are investigating with many signers! Still, sometimes you may only have a gut feeling of who would be good informants.
[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
You will want to think before you ask informants for help what you will be able to offer them in terms of payment; there may be standards in your university or institute for this. Also, it is a good idea to discuss before you make recordings on how you will use the recordings: request permission to show the video recordings to others during lectures at conferences or when teaching students, for example, if that is actually what you would like to do in the future.
If you want to gather spontaneaous data or elicit data, make sure that your informants are comfortable, especially when your informants are not used to having their utterances be videotaped. You can let them sign to someone to whom they sign more often, and you could try not to let the cameras be to prominent.
This may avoid:
In case you are not a fluent signer yourself, it is adviseable to ask a deaf collegue or deaf assistant to collect the data, in order to be sure to get the data that you want and to minimize influence from a spoken language.
[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
In some cases it is necessary to use signer's intuitions; this can be before you look at spontaneous or elicited data, so that you know what to look for, or after having looked at other data, in order to get more information.
Using signers intuitions (judgements) can be very tricky, especially if you are not a fluent signer yourself. Beware of the fact that subjects often try to satisfy the researcher in some way. Try not to tell your informants what you are looking for. Deaf people can be very good at reading people's expressions and may give you the answers that you (even unconsciously) want to hear, even if these are not (fully) truthful.
Asking for other people's intuitions can give a lot of information, but the interpretation is very difficult. If you are a late learner (and the knowledge of the sign language is still very restricted), you may not be aware of many aspects in the utterance under investigation that are needed for what you want to know, or that are present without you meaning them to be. This may blur the information or even put you on the wrong track.
It is not adviseable to ask informants for information on particular structures (e.g. "Do you think this sign is a compound or not?"), especially if they are not trained linguists. The man in the street is usually not aware of the linguistic structures of his utterances and his answers thus are not to be relied upon without further checking.
[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
Transcription of sign language data is a difficult and time-consuming process. Do not take it lightly, because it is very important! Before starting, one has to make a choice of what aspects should be transcribed and how. There are several ways and tools to use for transcription. Do not use a mere glos transcription, because this will give you far too little information about the sign language structures, and there is a danger that analysis has already started during the stage of transcription instead of after that.
There are several transcription tools and transcription standards that can be used.
Further information on transcription can be found at:
[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
Spoken language can influence your data. There are two types of influence:
- Processes of the spoken language are taken over in the sign language.
- Signers can use a form of speech supported sign instead of the sign language.
It is usually very difficult to find the influence.
In the latter case, in sign languages that are used in countries with a strong oral tradition the use of much mouthings of the spoken language during the signing can be an indication of spoken language influence.
[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
In presenting your work, either in presentations or publications, there are several ways to represent the data. Often a gloss notation is used, because it saves time and work. (reference to gloss notation systems?) In case you use a gloss notation, use glosses of the conference language or the language in which you write your publication; glosses in another language are usually incomprehensible by the audience).
However, note that many people don't know the language you are investigating, so a gloss representation is not always informative. In many cases it is also possible to use videos. You can also choose to use pictures or a notation system (such as SignWriting). A programme in which it is possible to quickly draw sign pictures is SignPS.
In order to facilitate comparison of sign (and spoken) languages or even make it possible, it would be nice if you could make your data available to other researchers.
[top] [data collection] [working with informants] [intuitions] [transcription] [analysis] [presentation] [references]
Also, the European project InterSign was partly focused on methodological issues; the web site contains abstracts of presentations and some online papers.
Baker, A., B. van den Bogaerde, J. Coerts & B. Woll (2000) Methods and procedures in sign language acquisition studies. InterSign Workshop, online paper.
Neidle, C. et al. 2000. The Syntax of American Sign Language. Functional Categories and Hierarchical Structure. Cambridge, MA: MIT Press. Chapter 2: Methodological Considerations.
Version: 14 April 2005