I've long been intrigued by the Friday night chatzi kaddish. Everyone seems to agree on the rough shape of the melody, but there are some obvious regional differences. Typically in the US, the melody for the "y'hei sh'mei raba" is the same as the one for the full kaddish at the end of the service, though that is usually in the major, where the chatzi kaddish is in the minor. This abrupt change in tonality has always seemed rather jarring to me. By comparison, UK congregations typically sing the "ba'agala" motif again for the words "y'hei sh'mei raba". Where did these differences come from? Is the American version a modern mistake? Across the 16 or so scores of this kaddish I looked at, one thing became abundantly clear to me. There is no authoritative version, and the only part of the kaddish which can realistically be called traditional or misinai is the "ba'agala" motif. Each composer has constructed a larger piece around this in a minor key, but they vary significantly. In many congregations nowadays, the "ba'agala" motif is used for the previous line "b'chayeichon" – this appears not to have been done historically. I found only one example of this; in all the other scores the melody starts at "ba'agala". When the motif reoccurs at the end, this is at "tushb'chata" in about half of the scores – the others start it only at "di amiran". The motif itself is sometimes written as a waltz, while some arranged it in 4/4. I have not seen anyone else mention this, but it is clear to me that it is the same melody as Eliyahu Hanavi which many people sing after Havdala (and is always in waltz time). It is not obvious what the thematic connection would be here. The motif has also been employed in other melodies: Naumbourg uses it in his V'sham'ru, and Baer uses it in his Ki Hinei Kachomer. Returning to my original question, what do these various scores include for the "y'hei sh'mei raba"? At least four of them use the full kaddish melody, but in the minor. So the American version is certainly not a modern mistake, as this usage is at least 150 years old. However I'd argue what we typically encounter in congregations today is unison singing on a harmony, not the original melody line, and because of this, the minor tonality is obscured to the extent that it sounds like we are modulating to the relative major. Abraham Baer's score is a fascinating departure because he intentionally modulates to the parallel major for "v'imru amen", and then uses the full kaddish melody for "y'hei sh'mei" in the major (but again this isn't the same as contemporary practice – the line is a third apart from what congregations do today). At the end of the kaddish, he performs the same modulation to the parallel major again for "v'imru amen", to leave us on a major tonality going into the silent amida. The only sources which use the "ba'agala" motif for "y'hei sh'mei raba" are the Blue Book and Marcus Hast, both published in the UK. Other scores which use the misinai "ba'agala" simply have their own versions of "y'hei sh'mei raba" which bear no resemblance to either the aforementioned motif or to the full kaddish melody. Many of these 19th century composers also include other alternative, unrelated tunes for the Friday night kaddish, and indeed there are some who only include non-traditional tunes. Many of these are in major tonality (e.g. Sulzer, Baer). Despite the misinai motif, there is nothing mandatory about using this traditional melody every week for this kaddish, and there have certainly been many other melodies written for it. Finally, there is a common practice to use this kaddish melody for the kaddish following the Torah reading on Shabbat day. How this became established is unclear to me, especially as other kaddishes have their own distinctive, unique melodies (c.f. festivals or Rosh Hashana).