Fieldwork update number 1

So, it's been ages since my last blogging, admittedly. The explanation is that I've been working hard for the past year on preparing for a data collection trip - December in Jerusalem, January in Malta.

The summer and fall semesters were spent on experimental design, item design, etc. Our incredibly patient and fun to work with native speaker of Maltese arrived in August to spend two weeks with us recording Maltese items - poor guy had to pronounce every Semitic verb in Maltese, plus about 10,000 nonwords in Maltese. We had planned to begin the same with a Hebrew speaker in September, but the original speaker bailed on us at the last minute, sending us into quite the panic as we searched for a suitable replacement. We finally found him, and he was also incredibly patient with us (he probably thinks we're crazy).

Amy and I arrived in Jerusalem on December 1, planning to run two experiments in the Verbal Processing Lab that Prof. Ram Frost directs at the Hebrew University of Jerusalem. Upon arriving, we encountered our first of several unexpected pitfalls (OK, we expect that there will be pitfalls, but we're never quite precise in advance about what exactly they will involve...)

The first pitfall involved our sound files. When we record our items in the recording booth in Arizona, we burn about 50-60 minutes of recording at a time to a CD, then upload the file to a computer, and our team of hard-working undergraduates goes to work in Praat adding a textgrid to the sound file, marking the best of the three tokens for each item with a transcription in IPA. We then run a super-cool Praat script on the textgrid that rips each transcription-item pairing out of the sound file into its own individual wave file. Unfortunately, for some reason our textgrid transcriptions involved a lot of unexpected carriage returns, which flabbergasted our script that otherwise works so well. Thanks to the astuteness of the script's author, Adam Baker, as well as valuable expertise from Paul Boersma and Jeff Berry, we were able to figure out what the problem was. And thanks to Jeff Berry's super powers, he was able to fix these textgrids so as to eliminate the unwanted carriage returns, and while he was at it, he avoided another potential pitfall by eliminating funny Unicode characters that our experiment software doesn't like.

This all set us back about a week. Next time, you'll get to hear about pitfall number 2...