From what has been said of the nature of dialect studies it is clear that some of them necessitate setting up an investigation project on an extensive scale. A lexicographical study of the whole language culminating in compiling a dialect dictionary or a comparative study covering the whole country with the intention of publishing a dialect atlas requires an active programme of data collection.
These investigations will produce manuscript material, usually in the form of phonetic transcriptions of spoken forms, written up in answer books designed to correlate with a linguistic questionnaire or vocabulary items taken down on index cards. Where sound recording is part of the project there will also be tapes of interviews in the field, incorporating questionnaire material and a corpus of uninterrupted speech.
Such linguistic projects, therefore, consist of a number of operations: planning the survey, field collecting and archiving of the data. After their completion the resulting studies can get under way. In setting up a project, however, certain basic requirements are necessary and decisions affecting the whole investigation have to be taken to ensure good order. These will be considered now.
In any undertaking which involves the planning and collecting of data in the field, the building up of an archive for the future and the promulgation of research, a centre where all the activities are coordinated is essential. At such a base the work can be organized and supervised. To ensure continuity it will also have the task of providing funds for its projects. Other research specialists and interests can be concerned with formulating policies but they should not interfere with the day-ta-day running. The Lexicography Foundation of the University of Helsinki, for example, 'was established by certain learned societies ... It was important that the Ministry of Education be included among the founders and the government was therefore persuaded to take part in the comprehensive undertaking ... Parliament awarded funds for word collection and when the continuation of state aid was promised it was possible to organize the work on a new basis' .21 But it is the Foundation which controls the project.
The centre will provide a base for staff, will house the necessary equipment for fieldwork and analysis of the speech-material, will provide conditioned storage rooms for the recorded audio-magnetic tapes and space for the catalogues, information retrieval systems and other archive material.
At a centre planning linguistic investigations a nucleus of staff covering different aspects of the work will be required. These will include:
staff to undertake collecting in the field;
technical help to ensure effective maintenance of the sound recording and auditory equipment;
archive staff to catalogue and arrange the collected material.
Where publication of the results is to follow, the aid of specialists - such as cartographers in making dialect maps - will be necessary. A research co-ordinator who will be concerned with planning what is to be investigated and who will supervise the programme of collection will be responsible for the complete undertaking. He should be able to consult specialists in the language in deciding the scope of the investigation, what it should include and its territorial extent.
The nature of the dialect investigation being undertaken will determine what kind of staff can be used for collecting in the field. The financial resources will dictate their number and how they should be remunerated.
To take down speech forms in transcription requires rigorous training in phonetics and, where a team of transcribers is employed in the field, even extra training is necessary in order to standardize individual practices in denoting shades of sound, in recording quantity and stress, in defining meanings of words and other matters. This was attempted in the case of the Linguistic Atlas of New England, as some nine fieldworkers helped with the investigation. Both for descriptive and comparative studies training in linguistics is also essential. As the research deals specifically with linguistic patterning the investigators should have a grounding in phonology and grammar.
In lexicographical projects the number of collectors is on a far greater scale, and it would prove too costly to take them all on staff permanently. In such projects they may be of two kinds. The trained collectors undertaking the main thrust of the investigation are usually people who have taken a degree in a linguistic subject. Their collecting can be financed by means of scholarships awarded for three or four years, so that they could collect in the same area for that period, and the total project can be managed both financially and territorially in this way. They will require training before venturing into a selected area. Some centres organize courses in word collecting so that such fieldworkers set about their task knowing what is required by the collecting archive if it is to have data that can be relied upon.
The second kind are people who will aid the project motivated by a natural interest in language: the voluntary correspondents. Finland has achieved a remarkable success in nurturing such correspondents: ordinary farmers, housewives, foresters and schoolmasters.
By means of a special magazine: Sanastaja ('word collector'), they are guided and made aware of what could be collected and are encouraged in their collecting. Such is the interest in this activity that at one time up to 1000 native correspondents sent in word collections to the archive. They are not paid for the work but may be awarded book or money prizes. Being people of the area they have the distinct advantage of knowing what local words, turns of phrase and idioms are used in the multitude of circumstances that make up human experience.
The activities of these two kinds of collectors complement each other. The trained collectors probe and elicit word forms, the voluntary collectors are awakened to be aware of the richness of their language and to offer from their vast repository of knowledge.
In all linguistic investigations fieldwork is a central activity and the necessary equipment to conduct it effectively must be procured. The collected data needs to be studied in more detail and suitable equipment is also required for this aspect of the work.
For fieldwork, both transportation and recording equipment need to be considered. The dialect fieldworker must be able to visit all parts of the country - even those not easily accessible - in the course of finding and choosing informants. For this travelling a suitable vehicle is a necessity if he is to complete interviews and recordings within a reasonable time. It need not be much more than a small robust vehicle able to carry the fieldworker and his equipment to his place of investigation. Under more arduous conditions a vehicle able to withstand rough road tracks and inclement weather would have to be used. For more sophisticated projects a recording car properly fitted for the purpose is often employed. Such an arrangement, however, calls for the services of a recording technician as well. 22
As the recorded sound material is subjected to painstaking auditory analysis and will in most instances be deposited in a sound archive for future consultation, it should be the aim to make the audio-magnetic recordings of the highest standard, bearing in mind of course the field conditions under which the interviewing takes place. Professional tape recorders, reel-to-reel, should be used where possible as domestic equipment is hardly suitable, and great care should be taken in choosing the right machine. The technical requirements for such equipment are further considered in the appropriate chapter of this publication.
As the fieldworker is at the same time the interviewer and the recordist, a high quality portable recorder which can be operated where there is no electricity supply, if needs be, should be chosen. Consideration must be given to microphones which match the recorder. Two high quality condenser microphones of compact size, one for the interviewer and one for the informant, are at present the best means of coping with recording two persons. Such a microphone can easily be clipped on to an article of clothing and arranged at an appropriate distance from the speaker's mouth. The voice level for interviewer and informant can then be adjusted separately.
An alternative method is to use a cardioid response dynamic microphone positioned on a boom arm stand. The microphone can be placed between interviewer and informant, again at an angle taking into account the voice levels of each.
Stereo recordings are far more difficult to complete and require specialist expertise. In spite of the great strides made in developing compact cassette recorders they do not appear as yet to equal reel-to-reel recordings in performance.
Additional equipment will have to be acquired for use at the archive. Here, listening in detail to field recordings in order to analyze the speech sounds of each variety and prepare phonetic transcriptions calls for a playback system that reproduces faithfully and clearly what has been recorded. It should comprise a tape recorder deck that can withstand periods of intensive use in replaying portions of tape, a good amplifier system and a separate loudspeaker. No erase head is required on the tape recorder since it might cause accidental erasures. Greater clarity in reproducing a recording at any required sound level can often be achieved by using a pair of headphones and their use has the advantage of not disturbing other people in the vicinity to the same degree.
For preparing copies of tape recordings other tape recorders are required and where funds allow the purchase of studio equipment should be considered.
A great deal of time is taken up at the archive in listening to the contents of tapes and to particular sound qualities within the 'utterances' (i.e. the complete response offered by the informant) so as to prepare accurate transcriptions. Two other pieces of equipment are useful aids in this listening process. The one is a 'tape-repeater' by means of which parts of an utterance can be selected from the interview tape and re-recorded on to a tape loop on another tape-deck. As its name suggests, this recording can then be repeated any number of times without having to replay the interview tape interminably. The complete section or parts of it can be concentrated on in greater detail to further the analysis. An even more specific portion of speech can be analyzed by means of the 'speech segmentator'. As before a portion of the interview tape is re-recorded and any brief part of the speech event - the onset to a particular consonant, the offglide of a vowel - can be isolated for repetitive listening and closer analysis.
Collecting in the field will result in the creation of manuscript and recorded material, to be sent back to the archive centre holding all the collected data. This data will obviously be used by researchers working on the immediate projects being undertaken, but it must be borne in mind that both the manuscript and the sound recorded material can serve the needs of different researchers and that other scholars may wish to consult it from time to time. As linguistic surveys are costly, further investigations of a similar kind are unlikely to be undertaken again for some time. Consequently if the collected data is to be readily usable, certain archival considerations have to be heeded before fieldwork has begun and the data, as it accrues, must be kept in a properly organized archive. The magnetic sound recordings should, of course, be stored under suitable conditions in the archive.
An appropriately designed answer book is used for taking down transcriptions of the speech forms of the informants. It is arranged in such a way that each answer correlates with a numbered question in the dialect questionnaire which is to be used to collect the data. A completed answer book will contain the data of one informant, each item in it being his answer to a question and the whole representing a variety of speech from one area. An extra column is usually provided to note down any 'incidental' material; that is, any further expressions used in the community which exemplify a particular speech form or any similar words having the same meaning.
Biographical details of the informant are also noted down so as to identify a data sample and, from each locality investigated, a number of such answer books will be produced. This will be the working programme throughout the whole area being surveyed. From these answer books the data necessary for mapping work, for example, will be abstracted.
Vocabular items collected when undertaking fieldwork related to lexicographical projects should be entered directly on to index cards, all of a specified size. These will come together in the archive and will be arranged according to a manner of indexing previously established. The information on each card should be set out according to a prescribed form. Normally it will include the index word in the standard language; the spoken form, which may be in transcription; its meaning, with complete utterances to illustrate its use. The card also shows the region from which the information was obtained, the collector's name, the questionnaire number and date of investigation. Thus one card contains details of a single meaning given to a word. All the fieldworkers, including voluntary collectors, should be taught to note down lexical items in this way, as this procedure will reduce considerably any recopying work which otherwise might have become necessary.
The vocabularies can be arranged in a single alphabetical index or in collections according to regions, as is required. With modern photo-copying devices copies of cards can be made so that different arrangements of the same material· is possible.
In section 4(c) of this chapter attention has already been drawn to the necessity of achieving the highest technical standards possible in field recording. Similar care and investment is needed in the archive to ensure the permanent preservation of recorded material. The main factors to be considered are the choice of tape, of tape formats and of recording speeds as well as the copying, packaging and storage of the archive collection. Regular conservation procedures should also be carried out. Readers should study Chapter 2 of this publication, and the associated references listed in Appendix A, for detailed advice on these subjects.