True sisters

The word for sister in Irish is deirfiúr /dʲɾʲəˈfˠuːɾˠ/, and it has always puzzled me why this word is so different from the words for sister in the other Gaelic languages: piuthar /pju.ər/ in Scottish Gaelic and shuyr /ʃuːr/ in Manx.

Yesterday I discovered that deirfiúr is in fact a combination of deirbh /dʲɾʲəv/ (true) and siúr /ʃuːɾˠ/ (sister). The word siúr originally meant sister in Old Irish, but came to mean kinswoman. To distinguish sisters from other female relations, deirb (true) was added to it, so the Old Irish word for sister was derbṡiur, which eventually became the Modern Irish deirfiúr – the s at the beginning of siur became f after mo (my), do (your) and a (his), and this mutation became fixed.

In Scottish Gaelic the word for sister came from Old Irish as fiur, which became piur and eventually piuthar.

The Old Irish word siur (sister) comes from the Proto-Celtic *swesūr, from the Proto-Indo-European *swésōr, which is the root for the word for sister in many European languages.

The Irish word for brother, deartháir /dʲɾʲəˈhaːɾʲ/, has a similar history: it is a combination of deirbh (true) and bráthair (brother) and used to be written dearbh-bhráthair or dearbhráthair. It comes from the Old Irish derbráthair, from the Proto-Celtic *brātīr, from the Proto-Indo-European *bʰréh₂tēr. In Modern Irish bráthair means brother as in a male member of a religious community or monk. In Old Irish it meant brother, kinsman or cousin.

Sources: Blas na Gàidhlig: The Practical Guide to Scottish Gaelic Pronunciation, by Michael Bauer
and Wiktionary

FacebookTwitterGoogle+Share
This entry was posted in English, Etymology, Irish, Language, Proto-Indo-European, Scottish Gaelic, Words and phrases.

10 Responses to True sisters

  1. joe mock says:

    What I find interesting is that both the ‘s’ and the ‘w’ of ‘*swesor survive, ‘s’ in the ‘sh’ of Manx shuyr and ‘w’ as the ‘p’ in Gaidhlig piuthar and the ‘f’ of deirfiur – but nowhere do they survive together.

  2. syntax says:

    Fascinating! Gura maith ‘ad as so!

  3. prase says:

    @joe mock: I wouldn’t say that the ‘p’ in piuthar descends from ‘w’ in *swesor. Assuming the evolution is described correctly here, it comes from ‘f’ in ‘fiur’, which is a result of mutation of the original ‘s’. On the other hand, the original ‘w’ could have contributed to ‘u’ in both Manx and Scottish versions.

  4. joe mock says:

    Are there other instances of s leniting to f/p?

  5. Andrew says:

    Interesting because even today in totally different cultures we refer to those who are ‘with’ us, part of our group or cause, as our “brothers” and “sisters”.

    Huh.

    Cheers,
    Andrew

  6. prase says:

    “Lenition” from s to f seems uncommon and I was unable to find other examples, in Irish nor in other languages. Specifically in “deirfiúr” I could hypothesize that first, siúr underwent the ordinary lenition to *shiúr [hu:r], leading to deirbhshiúr [dervhu:r] and then the consonant cluster [vh] by voicedness assimilation changed to [fh] and ultimately to [f]. Then the word, still considered compound, would be in Scotland reanalysed as “[der] + [f:ur]” which, stripped off the first part and lenition removed, would lead to the supposedly original [pu:r].

    But this is, unfortunately, only my speculation.

  7. Simon says:

    prase – you’re right about the process of change from s to f – I glossed over some of the details. The lenited form shiúr was re-analyised as being fhiúr, which is also pronounced [hu:r], and de-lenited to become fiúr. In Scottish Gaelic that form was thought to be the lenited form phiúr, which was de-leninted to piúr, and eventually became piuthar.

  8. joe mock says:

    Plausible, though it depends on ‘fh’ having been pronounced ‘h’ at some point – I suppose ‘fhuair’ would suggest that it was. Still, there must have been some distinction between ‘fh’ and ‘sh’ since the former disappeared and the latter remains. In any event, I’d be interested in seeing how i.e. *sw- developped in other words.

  9. prase says:

    I don’t think that it needs “fh” being pronounced as [h]. My hypothesis is that “bhsh”, pronounced [vh] turns into [fh] (not the Irish digraph “fh” but the sequence of phonemes [f] and [h]), which turns to [f] which is then reanalysed as “ph”, i.e. lenited [p].

  10. prase says:

    Simplification of consonant clusters on morpheme boundaries is not uncommon in Irish, cf. “tiocfaidh” where “cf” is pronounced [k] instead of orthographical [kf]. Unfortunately I can’t think of “bhsh” or “bhth” pronounced as [f] (or pronounced in whatever way, for that matter) – but that’s hardly surprising given my almost zero knowledge of Irish.