query: vocatives in modern Celtic languages

[Michael Daniel]

Dear all,

I am collecting data on vocative forms and forms of address, 
cross-linguistically, and am especially interested in languages that 
have special vocative morphology, or at least a vocative form that is 
distinct from the nominative. Starting from occasional references 
indicating the existence of the vocative in modern Celtic languages, I 
looked up a few sources. Being no Celtologist at all, I now seem to have 
a vague idea of the general picture now, but different sources are not 
always specific about all relevant points, so I would be very much 
grateful if you could correct me and update the gaps.

There is however quite a bunch of data mentioned and questions asked – 
takes some time to read this email, let alone to answer it. And also – 
you know how it happens, once you start answering one question…
Still, I am very much looking forward to all eventual answers!

Michael Daniel (Moscow)


My general understanding of the Modern Celtic vocatives.

Some Modern Celtic languages seem to have a vocative construction / 
morphology. The construction is particle “/a”/ triggering initial 
lenition of the following noun, or, alternatively, initial lenition 
only, with no particle, depending on the language. There are claims that 
vocative morphology (initial lenition) in the absence of the particle 
indicates that the particle has been lost (Ball&Fife 1993: 13), leaving 
lenition as a trace.

Specific languages.

Irish seems to have the “a”-particle-construction with lenition (e.g. 
Siadhail 1989: 149). Ibidem, page 149 he provides a special “vocative” 
form, which is not only lenited, but has a distinct flexion (‘man’ SgVoc 
/feara/ as opposed to SgNm /fear/; PlVoc/ feara/ as opposed to PlNom 
/fir/). I gather these forms are also only used after the particle a 
(although there is no direct indications at that). Very similar 
observations are made of Scottish Gaelig (Ball&Fife eds.: 1993: 173).

Welsh (King 1993, Ball&Fife eds. 1993) has a vocative lenition without 
the vocative particle. So does Manx (Ball&Fife eds 1993: 237). However, 
(Broderick 1984: 27) indicates that in what he calls ‘Late Spoken Manx’ 
only isolate vocative forms were used, he counts three nouns with 
attested vocatives in the singular and two in the plural. Some of them 
are exclamations (vuddee ‘O poor wench) rather than actual addresses.

Breton (Press 1986) does not seem to preserve any vocative lenition, 
whether triggered by a particle, as in Irish, or independent, as in Welsh.


My general questions are the following.

1. First of all, is all the data as interpreted by me and presented 
above correct?

2. Do I understand correctly, that in the languages that have a vocative 
particle the initial lenition of the following noun is not purely 
phonetic / phonological, but morphophonological (i.e. that it is the 
presence of the particle triggers it, not e.g. any element preposed to 
the noun and ending in a vowel).

3. May the vocative particle “a” in these languages be used 
independently (the way “hey” in English is, which can be used both 
independently and with a noun). From the descriptions I understand it 
can not, is that true?

4. Do the languages indeed fall into several clearcut groups – Manx & 
Welsh have no particle but have vocative lenition – Irish and Scottish 
Gaelig have lenition after the vocative particle – Breton has none. Or 
these are trends rather than rules (e.g. Welsh might occasionally use 
the particle, while Irish may sometimes drop preserving the lenition). 
Also, are flectionally distinct vocatives in Irish and Welsh _only_ used 
with the vocative particle? From the descriptions I understand the 
groups are clearcut – is that true?

5. Is the presence of a vocative morphologically distinct from the 
nominative (including e.g. Welsh where there is no particle and the 
lenition is to be considered as a grammatical marking) only 
characteristic of common nouns like ‘boy’, ‘child’, ‘god’ etc., as the 
examples seem to indicate, or are given names or family names also 
included into vocative morphology. (Mind that the example like /a 
Mha/ire/ in Siadhail 1989 is not distinct from the nominative, as far as 
I can understand, except for the lenition triggered by the presence of 
the vocative particle. I would be more interested in given names 
belonging to the ‘vocative’ declension type in Irish/Scottish Gaelig or 
proper names that lenite in Welsh / Manx). (King 1993: 19) notes for 
instance, that personal names in Modern Welsh are not subject to 
mutation – does that mean that forms of address with given names are 
always identical to the nominative in Welsh (and do they use any 
vocative particle then)? On the other hand, Scottish Gaelig names 
“Seumas” and “Donnchadh” belong to the relevant ‘A1’ declension type 
(Ball&Fife eds 1993: 173), so they are supposed to have distinct 
vocative forms, though no examples are provided. In general, is that not 
that the Celtic vocative, where available, tends to combine with common 
names rather than proper names such as given names? And what about 
kinship terms (especially ‘mother’, ‘father’)? Does it not tend to drift 
into exclamations (“Oh boy!”, “Good lord!”) from the pure vocative address?

6. Do I understand correctly that the /a /is possible in Irish and 
Scottish Gaelig not only with ‘1A’ declension type nouns (that have a 
distinct vocative form), but also with other nouns used as address. What 
are the semantic constraints on the combinability with /a /in different 
languages – only animates? Only humans?

7. How does the suggestion that the vocative lenition without the 
vocative particle is a trace of the vocative particle now lost combines 
with the fact that in a Manx example (Ball&Fife eds. 1993: 237) both the 
adjective and the head noun lenite? (If /a /is a true particle, it is 
very unlikely to repeat twice in one NP).

9. Finally, I am somewhat puzzled by a small note on Welsh (I understand 
it treats classic Welsh). (Jones 1930: 451) writes that “Some 
interjections are followed by nouns <…> Also, of course, by the 
vocative: /Och Dduw/. I wonder what he means here by ‘the vocative’. Is 
it a special vocative morphology? I was unable to find any mentioning of 
the vocatives in his grammar – what he does is that he explicitly 
indicates there are no case left in Welsh nouns.

For those of you who had patience to go through to the end of this 
letter – thank you, whether you think it is worthwhile correcting my 
mistakes and mis-interpretations or not.