The Recording Process
The recording devices used in the pilot study were the best of their kind at the time: Sony TCD-D100 DAT recorders with Sonic Studios stereophonic DSM-1S/L microphones. The recording equipment was put into a pouch to be carried on the belt or in a bag. The microphones were attached to a device that was carried on the informant’s neck, so that the microphones were located one at each respective side of the informant’s head, close to the ears, thus recording exactly what the volunteers heard. The cable connecting between the microphones and the recorder was hidden beneath the informant’s clothes (see illustration). This way, the recording hardware would not attract any attention of either interlocutors or the surrounding people. This, indeed, proved to be the case in most instances.We used 4-hour digital (DAT) Sony 120P tapes. Following the procedure that each informant would record between 8 to 16 hours of longitudinal recording, each volunteer would use two to four tapes. The tapes were later backed up on CD-ROMS. Each set of recordings that includes everything that was spoken during the 8, 12 or 16 hour period, apart from the short intervals that were needed for cassette switching.
Awareness of being recorded can influence one's speech usage. This awareness normally subsides after a short period of time. The impression one gets while listening to conversations dealing with the recording itself is that they usually do not use a form of speech that differs from the regular, spontaneous speech of the speakers.
A good example of this is a recording in which the volunteer, having arrived in her workplace in the morning, tells a coworker about the recording and asks her to provide some personal data (C514_1). Yet when a new topic is introduced, the circumstances of being recorded are forgotten and the conversation proceeds spontaneously and excitedly. Based on the remainder of this conversation as well as other recordings, one gets the impression that even if awareness of being recorded intermittently reappears, it lasts for only a short while. At any rate, it does not seem to impair the nature and character of the spoken language (cf., e.g., C714_sp1_029-032 and its later parts; C532:19”-80” [parallel to lines 4-45 in transcript C532ND]).
One other example comes from a recording that was made during the preparatory phase, in which the recorded volunteer had prior knowledge of the object of the study. This volunteer, "Omer", was requested by his father, a representaive of CoSIH, to augment his previous recordings by an additional 30 minute recording documenting a different genre. In this latter recording (OCh), "Omer" tells his father about his trip to China and Mongolia. At a certain moment during the conversation, he inquires about the recording length and is requested to continue his narrative. It seems that the language he uses when talking about the recording, as well as the language he uses later, do not deviate from the spontaneous language of the conversation in general (cf. OCh_sp1_807-812; sp2_244-245 and later). Conversely, at the beginning of another recording, the volunteer designated as C2 addresses his would-be listeners using language that is markedly different from his everyday language, discernible in his conversation with his wife and friends (cf. C211_1:17.50”-45.64” [parallel to lines 14-37 in transcript C211_1ND]). This volunteer is aware of the differences between the way he expressed himself while addressing these potentional listeners and the way in which he normally speaks, and even makes explicit mention of this fact.
We tried to overcome the problem of linguistic awareness using the following procedure: when informants were approached by our representatives, they were told that the goal of the research is “recording the daily life of Israeli inhabitants”. Although this is not the whole truth, it is the truth, and nothing but the truth. When our representatives came to collect the recordings and before working on the sociolinguistic questionnaire, then they told the informant that the recordings would be used for the compilation of CoSIH and asked for the informant’s consent to use the data (Izre’el & Rahav 2004: 4a; cf. יזרעאל תשס"ג: 216). Naturally, the question of awareness and language form deserves further study. References to the recording are found in a number of texts, some of which were transcribed by Nurit Dekel (see Table 3: Recordings, Transcripts and Transliterations). These may perhaps allow us to better investigate this issue.
All volunteers took part in a sociolinguistic interview at the end of the recording period allotted to them. They were asked about their social and linguistic background, as well as about other relevant matters. These data are at the disposal of the users of CoSIH. The informants and their interlocutors are identified only in pseudonyms. Their privacy is maintained both in the transcripts and in the recordings, in which all names were replaced by tone signals. All instances of names were cut from the sound file and converted to tone signals, which, in turn, replaced the discarded segments. This conversion procedure was made in Praat.2
2 In the Periodicity menu of the Objects window, the To Pitch command creates a tonal curve. This curve can be converted to sound using the "To Sound (hum)" command that is found in the same window. We thank Noam Amir for his advice on this procedure.