In the course of an explanation of site changes, Anggarrgoon presents a most interesting explanation of a bit of Bardi grammar:

“look like” or “resemble” is irrganbala, it’s a noun, it’s inalienably possessed, it’s (understandably) obligatorily plural. It is also one of very few words to take the “spouse” suffix –milj. I call it a “spouse” suffix because it marks “appropriate” pairs, e.g. iilamilj, a dog and its mate. irrganbalamilj, though, means “they look like each other”. I can’t remember an etymology, if I ever found one, but it bears a suspicious resemblance to the word for track, footprints, niinbil or niinbal (which would be what we’d expect for the singular, from ni-ganbala, which I think would have to be reconstructed as *niganbila to make the vowels turn out right.) Not to be confused with niyambal, niimbal ‘foot, footprints’. n doesn’t normally assimilate to b.

I have to say, much as I enjoy explaining the role of coincidence to people convinced two similar-sounding words must be related, even I find it hard to believe niinbal and niimbal, both meaning ‘footprints,’ are unrelated. Not saying it ain’t so, just pointing out that I’m not immune to the natural human craving to connect similar things.


  1. Ha ha.
    Resist, Hat, resist, and keep in mind that if “bell” and “belfry” are unrelated (which they are), and if “minimum” and “miniature” share no kinship (which they don’t), perhaps only a nimble trick might yoke “niinbal” to “niimbal.”
    More to the point, I have never heard of this tongue Bardi! Though, I must say, it sounds epic.

  2. Heh. Bardi — and looking at that Ethnologue article, I see the primary spelling is BAADI. Goddammit, is this another of those Brit spellings where -r- is used pointlessly to extend a previous vowel??

  3. Hi everyone,
    That ethnologue spelling is as right as Bardi, ie they’re both wrong. Phonemically, the language is /ba:d,i/ (long a, retroflex d). In the practical orthography, that comes out as Baardi. When the community orthography was codified in 1990 they decided to make the language name an exception to the spelling, because “Bardi” was on the store, clinic, school, community documents, etc etc. The ethnologue name comes from, I think, the Capell/Elkin/Worms/etc 1950s work, ultimately. They were ok on vowel length but had a lot of trouble with retroflection. I’ve pointed this out to Ethnologue, twice actually, but apparently getting the primary spelling changed is harder than the 3rd or 4th hardest thing you can think of.
    I should have also mentioned re niinbal and niimbal: if I’m right, the first comes from a stem -ganbal, the second from -jambal. The ni- is the 3rd person possessive prefix. ijV and igV both get lenited and come out the same. You can see the stem in “their feet” – irrjambal, where there’s no lenition.

  4. Thanks for the elaboration! (It must be pretty cool to see your name attached to an Ethnologue article…)

  5. “is this another of those Brit spellings where -r- is used pointlessly to extend a previous vowel”
    No it’s an AUSTRALIAN spelling where -r- is used pointlessly to extend a previous vowel.
    Why I oughta…

  6. but it’s not pointlessly extending the previous vowel, it’s non-pointlessly indicating retroflection on the following consonant.

  7. Oops, I didn’t read carefully enough and was just taking umbrage at LH’s ‘British’ slur. But thinking about it, given that the retroflexion isn’t something that most Balanda are going to manage, for non-Bardi speakers the r is going to (actually non-pointlessly) lengthen the vowel, which is probably better than nothing, dontcha reckon?

  8. Sorry about the careless “Brit.” But the point still stands: for those of us (the majority of English speakers) who actually pronounce r’s, these spellings are actively misleading. They should be banned, I tell you! Wiped off the face of the earth!

  9. yeah, and “Bardi”, however it’s said, is still a good deal better than some of the things I’ve heard my friends called by white people. Actually, your writing Balanda (

  10. To a Swede, writing -rd- for a retroflex -d- is a rule, to which there is no exception in most dialects. It even sometimes(?) always(?) works across a morpheme boundary -r|d-.

  11. Lars Mathiesen (he/him/his) says

    Swedish: It does work over morpheme boundaries, even “traditional” word boundaries: har du sett = /hɑɖʏʷ’sɛtː/ vs du har sett /dʏʷhɑ’ʂɛtː/ or du har fått /dʏʷhɑrˈfɤtː/

    “Spouse” suffix: Danish uses magen til, originally ‘the mate of,’ for two identical things (up to symmetry, like shoes). Also mageløs = ‘peerless,’ this one in Sw too (makalös). But we don’t call a person magen til their spouse. (Ægtefælle is traditional, but people don’t get married so much any more. We use various words that compositionally don’t imply emotional bonds, like samlever, pragmatically distinct from ones that actually don’t, like bofælle ~ ‘flatmate’. Swedish has innovated here: sambo came first, for [unmarried] couples living together, and then särbo for couples maintaining separate households. [isär = ‘apart’]. Such arrangements don’t engender any legal advantages, and in Danish we’d have to resort to kæreste = ‘fiancé(e)’ [orig. = ‘dearest’ + zero nominalization]).

Speak Your Mind