o-vocatives: Analogical account, Part III

In the last few posts, I’ve worked through the analogies that have extended the o-vocative into proper names: M1–M5, O2–O6. There was to-ing and fro-ing, there was nebulous definition and redefinition of rules, there was a whole ballet of criteria. But the ballet orchestration can be formulated: the rules for the analogy are sweeping, even when they seem to be contradictory, and could be distilled in a handful of features: NAME, VERNACULAR, FAMILIAR, PENULT-STRESSED. (Those features are at work even when the rules were redefined: the diminutive -akos suffix for surnames in 1941 was an analogy based on the FAMILIAR feature; the etymological transparency criterion used now is an analogy based on the NAME feature—surnames that look like common nouns are less NAME-like than opaque surnames, so they tend to take the o-vocative less.)

That leaves us with the common nouns of O1. The story for those is going to be less satisfactory, because the analogy does not involve entire well-defined classes of name: they are more typical of analogical spread, in that they involve individual words, and the commonalities at play are harder to discern. This is an area where I may need to be enlightened by Henrich’s thesis on the o-vocative, once I do read it.

The one exception documented in Triantafyllidis’ 1941 grammar, for example, is O1a: kapeˈtanios “captain”, kamaˈrotos “porter, cabin body”. Why the navy would be fertile ground for o-vocatives early on, to the extent of extending them to common nouns, is not obvious. They are both Venetian words, and their foreign origin might have encouraged some sort of grammatical streamlining (like I surmised for final-accented names like nikoˈljo < Niccolò); but for those two naval terms, there are dozens of Venetian nouns like kuˈnjaðos “brother-in-law” or maraˈnɡos “carpenter”, which form their vocatives regularly.

The temptation is always there, whenever maritime vocabulary is involved in the Mediterranean, to invoke the Mediterranean Lingua Franca: maybe the o-vocative with maritime vocabulary reflects some sort of pidginisation of Greek, that would have been all too happy to dispense with its now anomalous e-vocative. That’s a little too speculative for my liking, though.

It’s safer to posit that the source of all the analogical extensions of the o-vocative, ˈɣeros “old man”, extended to random other honorifics, one of them being “captain”, and that “cabin boy” patterned after “captain” because it was another naval term.

That’s safer. It’s not more satisfactory, I’ll grant you.

Commenter Pepe had an intriguing hypothesis: “captain” and “old man” are both routinely prefixed before given names (ˈɣero-ˈnikos “old man Nick”, kapeˈtan-ˈnikos “captain Nick”). The prefixed form of “captain” is truncated, but he thinks the o-vocative appears with those common nouns in isolation, because they are so strongly associated with given names. We have seen that historically ˈɣero probably came first (at least, Henrich’s third declension theory explains the parallel ðjako, ðrako “deacon, ogre”); but the association Pepe posits is real enough. I hesitate over it because it requires analogy to apply even when the name is absent, and to generalise from truncated kapeˈtan to the complete vocative kapeˈtanjo. But it’s not impossible.

If there were a foreign-language bias towards the o-vocative, that would also explain it showing up in kumˈbaros “god-sibling”, which is also Venetian. (147 hits for re kumˈbaro on Google, 14.4k for re kumˈbare.) But the explanation is likelier that ˈɣeros “old man” brought it about, as an extension of honorifics to familial terms.

In fact, the most controversial extension on the thread in Sarantakos’ blog was ˈθios “uncle”. There was much discussion about whether it was a regionalism, whether it was slang, and what its precise semantics was with an o-vocative. As Evangelos Lolos insightfully pointed out, there is nothing rural about ˈθios: with its V-V hiatus, it is clearly a learnèd form, and the traditional vernacular forms are either θjos (which forms its vocative regularly as θje), or the Italian loan barbas.

But the learnèd form has been pressed into service in a vernacular function: “uncle” is also used traditionally as an honorific for any older stranger. (That survives jocularly in barbas, but of course the etymology of the word, “beard”, shows that it did not start out as a joke.) My own experience is that if anything, ˈθios is likelier than θjos to be used as a title of respect: it is after all a more polite-sounding word, being phonologically learnèd. And as a title of respect towards elders, ˈθios was more susceptible to influence from ˈɣeros “old man”.

If ˈθios is slang (μάγκικο), and less likely to be used with actual relatives, as it is for Sarantakos, that is all the more reason for the analogy with ˈɣeros to take hold. But this is messy enough that commenter Maria claims the opposite pattern, with the o-vocative used for the relative, not for the honorific directly analogous to ˈɣeros:

If I addressed my uncles as ˈθie, or older men in the neighbourhood as kirie “sir” instead of ˈθio, they would have thought I considered them strangers. I only addressed one uncle as ˈθie, after being duly tutored. But he was a judge.

The miscellaneous terms in O1c are harder to explain because they are miscellaneous; but I suspect they are all tied together by the feature FAMILIAR, which we have already seen at play in given names, and which in common nouns is at its most pronounced when it is used as an overt signal of disrespect.

The miscellany are:

• ˈɣiftos “Gypsy, blacksmith”
• kaˈkurɣos “criminal”
• ˈbufos “horned owl; fool”
• ˈmurɣos “shepherd’s dog; uncouth man”
• slang ˈfilo “buddy!”
• military slang ˈneo “newbie!”

In the 1973 song Come, gypsy/blacksmith, gather hammer and anvil, which I have already posted on here, there is nothing necessarily racist going on: blacksmiths were traditionally Roma, and the song only refers to Roma in that capacity. But of course, there has been no shortage of contempt from ethnic Greeks towards the Roma, and the o-vocative is consistent with that. A word used as a slur (even if it is the unmarked traditional term for Roma) is going to choose the familiar-coloured vocative over the default vocative. That also applies to the other instances of invective on the list, ˈbufos, ˈmurɣos, and the surprising entry in the list, kaˈkurɣos, which ostensibly looks too learnèd to tolerate an o-vocative.

Contemptuous familiarity, for that matter, is overtly behind the military slang ˈneo, as Pepe remarked explicitly; and it’s not that far away either from ˈfilo (Pepe’s example in the same comment is τι μας λες ρε φίλο; “what are you trying to tell us, buddy?”) In fact, we have already seen contempt in the diminutive anθroˈpako “puny little man” (although its parallel filaˈrako “buddy” is amicable, not contemptuous, in its familiarity.)

So it is possible, just, to discern some unifying trends in the few common noun instances to date of the o-vocative—other than the inherited constraint that they are all penult-accented:

• There’s the initial category of third declension nouns (M1);
• Two maritime terms (possibly as a one-off extension of ˈɣeros);
• Two familial terms (possibly as another one-off extension of ˈɣeros, with the secondary sense of ˈθios “uncle” in play as an honorific towards elders);
• Terms likely to be used as disparaging terms, either because of their literal content (“criminal, fool, uncouth”), because of racism (“Gypsy”), or because of slang register (ˈfilo, ˈneo).

The military slang of ˈneo, by the way, can be interpreted in a different contemptuous way: as a neuter, which also ends in -o. The neuter with reference to humans in Greek has a connotation of infantilisation (since children are pre-sexual). It can be used as a sign of affection; and indeed, as Pepe had pointed out and I had failed to acknowledge, the neuter could in fact be behind the putative masculine final-stressed given names in M5, like manoˈljos, nikoˈljos: manoˈljo, nikoˈljo could equally be the vocatives of the neuter hypocoristics manoˈljo, nikoˈljo. But the infantilising neuter can also be used dismissively, particularly in a hypermasculine domain like the army.

There is a different variant vocative which also comes from the military, and which is also used negatively: the use of the nominative instead of a vocative. This extends to first as well as second declension, and is independent of accentuation. In the military, it is de rigeur with surnames, and is how orders and punishments are barked at soldiers: Παπαδόπουλος! Δέκα μέρες κράτηση! “Papadopoulos! Ten days confined to quarters!” The point of the nominative is that the army grunt should feel talked at, not talked to, as part of breaking down their sense of individuality. (What Wikipedia politely refers to as resocialisation.) The military nominative-as-vocative turns up now and again in more general slang, most notably in the not particularly respectful ˈkirios “hey Mister!” As elucidated at slang.gr:

Form of address towards men, usually strangers, not particularly polite nor outright rude. It has long been used to express restrained irritation, combined wiht distance. It is not so familiar as to imply a provocation to fight, but it won’t say no to a few cross words. The counterpart for women is maˈdam.

Nowadays ˈkirios is even used by students towards their high school teacher, without necessarily being provocative. By analogy, the vocative ˈfilos, which can be completely decorous and friendly. [The corresponding entry in slang.gr notes that it can be “either negative or positive”.]

Note: the use of the nominative instead of the vocative, which can only happen grammatically with masculines, expresses distancing. It is used routinely in the military.