24 November 2015

Boontling “Deek”: A Rovin’ Gypsy Word?

The little town of Boonville (Mendocino County, California) was established in the early 1860s near a slightly older place called The Corners. A local general store was moved from The Corners to the present location of the town centre and then sold to Mr W. W. Boone, who modestly named the settlement after himself (it had briefly been called Kendall City in appreciation of another local businessman). The inhabitants of Boonville (now about 1000 people) refer to their town colloquially as Boont.

What makes Boonville special is its local ‘jargon’ which probably arose in the 1890s among children and young people (who then grew up without abandoning it). The community was quite isolated at the time, and kept no records to inform posterity why they chose to develop an extremely hermetic and highly inventive vocabulary of about 1500 words, known as Boontling (Btl.). Boontling was not originally meant to be written down, but a semi-formalised spelling was developed for it in the 1970s. One of the local words is to boont ‘to speak Boontling’. At present Boontling is dying out (Btl. pikin to the dusties) despite having been discovered by linguists and made known to the general public. Many Boontling words remain in circulation, but there are few fluent users left. Boontling has never been a fully fledged dialect: it has a distinct vocabulary incomprehensible to outsiders, but the accent is a rural variety of Northern California English (with historical affinities to the Midwestern and Border South dialects), and Boontling syntax is in nearly all respects the same as that of mainstream US English.

The Old Machine Boys [source]
Despite its recent origin, Boontling vocabulary is etymologically opaque to a surprising extent. Nevertheless, the vast majority of its words are coined from pre-existing elements rather than made up entirely from scratch. Often you have to know the history of the place and rely on anecdotes collected from elderly locals that “explain” the meaning of some words, especially those derived from personal names. (A professional etymologist has to verify their historicity, of course, and this is likely to be the toughest part of the job.) Some words reflect otherwise forgotten dialectal or slangy vocabulary. Some were coined using Humpty Dumpty’s technique of piecing together broken fragments of ordinary English words. Some hide behind strange pronunciations that appear to have been borrowed from Scottish or Ulster Scots speakers. Some came from Spanish (approximately half the population is of “Hispanic or Latino” descent), and a few from the Pomoan languages indigenous to California (there are a few Native Americans as well).

I’m intrigued by a few of them. For example, one of the most common and persistent Boontling words is deek ‘look, see, stare, notice’ (also used as a deverbal noun). I’m not aware of the use of deek anywhere else in North America. However, deek is a well-known colloquial Northernism in Britain. It’s stereotypically associated with Geordie (the dialect of Newcastle and the Tyneside area), but it actually occurs throughout Northern England (including Cumbria, Liverpool and Yorkshire) and much of Scotland. The word is a loan from Romani or rather Angloromani – the Romani-derived lexicon embedded in the varieties of English used by the British Romanies (see Yaron Matras, 2010, Romani in Britain: The Afterlife of a Language, Edinburgh University Press). The Angloromani verb (no longer inflected) is deek, dik, dikkai [diːk, dɪk, dɪkʰaɪ], reflecting European Romani dikh- ‘see’. There are, by the way, quite a few Romani loans in British dialects (some of them, such as pal ‘brother, friend’, no longer dialectal). The Dictionary of the Scots Language gives, among others, these recent examples of the use of deek:
  • Deek that gadgie. ‘Look at that guy.’ (Edinburgh, 1988)
  • The gaffer wis anither big rough-deeking gadgie... (Aberdeen, 1990)
Here, in addition to deek, also gadgie guy, bloke is a Romani loan (Angloromani gadji, gawdjo, gawdja < European Romani gadžo ‘non-Gypsy’).

The root dikh- arrived with the ancestors of the modern Romani all the way from Northwestern India. It is cognate to Hindi dekh- and to Sanskrit dṛś-, dṛkṣ-, all of which continue a well-known Proto-Indo-European root, *derḱ- ‘watch, see’. Incidentally, the Hindi word became independently borrowed into British English via the army slang of British soldiers serving in India, hence have a dekko ‘have a look’.

The Germanic languages also inherited a few words derived from *derḱ-, but English has lost all of them. Old English still had torht ‘bright, splendid, illustrious’ from the PIE deverbal adjective *dr̥ḱ-tó- (cf. Skt. dṛṣṭá- ‘seen, visible’). It was used almost exclusively in poetry, but also served as an element forming personal names. For example, an Old English gadgie called Torhthelm (Totta for friends) owned a farm called Totta’s Homestead (Tottan-hām) in todays north London. The To- part of Tottenham is about all that has survived of the root *derḱ- in Modern English via direct descent. A number of other reflexes, however, have reached English by horizontal transfer from other Indo-European languages, the most spectacular of them being dragon (ultimately from Greek drákōn ‘starer’ → ‘serpent with a deadly stare’). But I’m digressing.

I have no watertight proof that Btl. deek is the same word as Angloromani, Northern English, Scots and Scottish English deek, but I’d be very surprised if somebody proved that Btl. deek had a different origin. Still, I have no idea how the word could have reached an obscure valley in Northern California and become fixed in the local slang without leaving any other traces in American English. If anyone among my readers comes up with an idea how to explain its trajectory in time and space, I’ll be immensely grateful for sharing it.

42 comments:

  1. Drakon = starer. Sounds interesting. Greek mythical drakones were simply giant snakes, but this meaning would imply terrible stares as basilisks or Medusa?

    ReplyDelete
  2. This is the usual (anf formally umproblematic) etymology, accepted also by Beekes in his Etymological Dictionary of Greek. Yes, it would mean that a paralysing/killing gaze was attributed already to (pre-)Greek dragons. An alternative possibility is 'watcher' (a guarding serpent), like Ladon in the Garden of the Herperides.

    ReplyDelete
    Replies
    1. and ... unproblematic ... Hesperides

      Gosh. After-midnight typing has a strong mutagenic effect.

      Delete
  3. One syntactic difference is known, and appears in this 1980 [sic] sentence: "You must do much graymatterin fore pikin for seekin Ite steaks to gorm, cause the sockers might not be bahlers, but nonchers with dusties dust, so deek your bok well", where "pikin for seekin Ite steaks" = "going to find mushrooms" in Standard English.

    Also, moshe is < machine, but it means 'car, auto', so the caption means 'The Old Cars Club".

    ReplyDelete
  4. I had pikin for in mind when I wrote the syntax was "virtually" the same.

    Moshe can actually mean any "device", not even necessarily mechanical, cf. harpin moshe 'radio'. The county fair exhbition was of tractors and other farming equipment, so I thought "Old Cars Club" would be misleading.

    ReplyDelete
    Replies
    1. Oops, not "virtually" I changed that to "in nearly all respects".

      Delete
  5. You might find this of interest: http://www.languagesoftheworld.info/the-americas/language-dialect-or-jargon.html

    ReplyDelete
    Replies
    1. Thanks, Asya. I of course agree that Boontling is a jargon or cryptolect rather than a separate variety of English (let alone a language). My point here is that even if people deliberately create an "encoded" register for in-group communication, they usually recycle and modify something already available, The process is not much different from ordinary lexical innovation, only accelerated in a small and dense communicative network.

      Some secret languages have their own phonology and "orphan" roots for which no cognates have been found anywhere. Damin, formerly spoken on two islands in the Gulf of Carpentaria, is like that. It has a unique sound system with, among others, the only inventory of clicks found outside Africa, an ingressive lateral fricative, a bilabial trill, a velat ejective, and a bilabial-velaric eggressive (a kind of inverse click, not found in any other human language). It also has only two pronouns, 'me' vs. 'other(s)' (as opposed to nineteen personal pronouns and several demonstratives in Lardil, the "normal" language of Damin users). But apart from its inventory of 150-250 "secret roots" with their own exotic phonotactics, the rest of the system, including derivational morphology and syntax, is basically that of Lardil, and affixes added to Damin roots are pronounced as in Lardil. So even in this extreme case the "secret language" is actually just a ceremonial register embedded in the host language.

      Delete
  6. Fascinating post.

    Is there any etymological relationship between 'gadžo' and 'goy'?

    ReplyDelete
    Replies
    1. The origin of gadžo is uncertain; there is no obvious Middle Indo-Aryan source. The most popular hypothesis is that it comes from Old Indic gārhya- 'domestic' (i.e. 'settled' as opposed to 'itinerant'). The adjective itself is a derivative of gṛhá- 'house, home' (thus already in the Rigveda), usually analysed as *gʰr̥dʰ-ó-, and regarded as cognate to Goth. gards 'house, courtyard, enclosure', Slavic *gordъ 'fort, town', etc. (also to yard, garden and gird).

      It is speculated that gārhya- became (unattested) Middle Indo-Aryan (and Proto-Romani) *gajjha-.

      As for goy, the word goes back to Biblical Hebrew gôy meaning simply 'a people, nation' (from Central Semitic *gāy‑ 'tribe'). Originally, it included the Israelites, but in more recent times came to refer exclusively to nations other that the Jews (i.e. the "gentiles"). There's no connection whatsoever with gadžo.

      Delete
    2. P.S. It would follow that gadžo is a distant relative of Russian graždánin 'citizen'.

      Delete
    3. usually analysed as *gʰr̥dʰ-ó-

      Not that I have a better idea... but is *dʰ > h regular? I can imagine this happening behind [r], but I've never heard of it.

      Delete
    4. The "debuccalisation" (loss of the oral articulatory component) of the reflexes of PIE *ǵʰ and palatalised *g⁽ʷ⁾ʰ is of course regular. But the same loss occasionally affected * and (later and more rarely) *. Have you never seen hitá as the past participle of the root {dhā} 'put, place' (*dʰeh₁-)? It's probably the best-known case, but there are many others. The relevant material is presented and discussed in this article by Sasha Lubotsky (1995). It isn't clear how much of * > h [ɦ] is due to sporadic or semi-regular change in Vedic itself and how much to interdialectal borrowing within Old Indic, but Lubotsky discussed that question as well. See also Kobayashi (2004), especially p. 84-86 (he calls the process "deocclusion").

      This kind of phonetic weakening gained momentum in Middle Indo-Aryan, but that's a separate story.

      Delete
    5. Have you never seen hitá as the past participle of the root {dhā} 'put, place' (*dʰeh₁-)?

      Heh. No – in linguistics I'm a complete and utter autodidact. In this day and age (thank Gore!) that gets me pretty far in some randomly selected respects; but I have never taken a course in linguistics, and have never tried to study Sanskrit (or even Greek), not even at the "grammatical sketch for comparative linguists" level.

      (The only part I can recite of all the Vedas is, well, this.)

      Thank you (and Gore!) for the papers you're linking to; I'll try not to stay up all night. :-)

      Delete
    6. No need to worry; as it is, you are probably the best linguist among biologists. Well-informed autodidacts are my favourite audience.

      As regards the Vedic fragment in question -- the Slovak translation is a real gem.

      Delete
    7. probably the best linguist among biologists

      Faint praise :-þ

      the Slovak translation is a real gem.

      Isn't it. :-)

      Delete
  7. So did Grassmann's law not affect laryngeals? Or is it just that it came into effect in Indo-Aryan after their loss?

    ReplyDelete
    Replies
    1. I don't think I understand the question. There was no PIE (or PIIr.) laryngeal in *gʰr̥dʰ-ó-, but since it contained two mediae aspiratae, Grassmann's Law affected it normally, yielding Proto-Indic *gr̥dʰ-á-, with the further weakening of *-dʰ- to Vedic -h- in this word. Its Iranian cognates preserve a stop articulation (Iranian *-d-).

      Laryngeals as such were not affected by Grassmann's Law in any way, but the aspirating effect of *h₂ on a preceding stop was older than Grassmann's Law and triggered its application. One of the most obvious cases is *dʰugh₂tér- 'daughter' > Proto-Indic *dugʰHɨtár- > Ved. duhitár- (measured with a heavy first syllable in the RV!).

      Delete
  8. Sorry I was referring to *dʰeh₁, but you've answered my question.

    ReplyDelete
    Replies
    1. I see. In my opinion (nor shared by everyone, but I think representing the majority consensus), only *h₂ had an aspirating effect in Indic (which doesn't mean that the other laryngeals were simply deleted: they survived long enough for their metrical effect to be visible in the RV). So if any of them could be suspected of having participated in Grassmann's Law (the way Proto-Greek *h < *s did), it would be *h₂ but there are roots like *bʰah₂- 'speak' > Ved. {bhā}, in which no deaspiration took place, so it seems *h₂ could trigger Grassmann's Law only indirectly, via secondarily aspirated stops.

      Delete
  9. Those metrical traces in the RV and Homer are tantalising. Whatever you call the languages that contained direct reflexes of laryngeals, they must surely have been pretty damn close to already being Vedic Sanskrit and Greek.

    ReplyDelete
    Replies
    1. Are you sure there are traces of segmental laryngeals in Homer's metre? Words like tʰugátēr or platús have a light initial syllable there, as opposed to Ved. duhitár- and rátha-. Even in Vedic, the loss may have happened earlier than assumed by Kuryłowicz, Schindler or Gippert (1996). The heavy scansion attributed to a consonantal laryngeal is not applied consistently in early Vedic, and, what's even worse, there are words in which it applied though there's no evidence of a laryngeal. Therefore, while the metrical licence to treat synchronically light syllables as heavy in some positions is probably a laryngeal effect, its actual distribution was already strongly disturbed by early Vedic times. Martin Kümmel, for one, is sceptical of dating the loss of laryngeals as late as Proto-Indic, though he too agrees that consonantal reflexes of at least *h₁ and *h₂ persisted well into common Proto-Iranian.

      Delete
  10. No, I'm far from sure - it's 20 years since I did historical linguistics! I'm returning to it mainly through your blog.

    I have a memory of there being evidence from Homer in James Clackson's textbook, but I'm probably wrong.

    ReplyDelete
  11. Or, judging from your italics, what I'm remembering is not metrical evidence.

    ReplyDelete
  12. Maybe what you're remembering wasn't about laryngeals, but about syllabic /r/. Indeed there's a place in the Iliad where a heavy syllable has to be read as light, which becomes possible if you reverse-engineer it to having a syllabic /r/. However, as discussed in French in the comments here, it seems to be the mainstream hypothesis nowadays that the archaic Greek epics weren't learned by rote – instead the singers learned the names of the acting parties and what they did in each line, and filled in the rest with stock phrases. If that is correct, the text as we have it cannot preserve much older features.

    ReplyDelete
  13. It wasn't in Clackson's book, it was in Winfred Lehmann's:

    "Greek metrics gives us further evidence that laryngeals formerly were found initially in certain words. Such words beginning with λ, μ, ν make position, e.g. νέφος. Assumption of an initial laryngeal is supported by cognates with /en-/, from /Xen-/, in other dialects; compare Skt. ámbhas, Gk. νέφος. (Lang. 17.91–2; LaRoche, HU 49–59 gives a complete list of such words.)

    ...

    From these irregularities we can determine laryngeals initially, medially, and finally in words. The irregularities are rare. We may assume that they were removed as the Vedic and Homeric poems were handed down. If we find them surviving, we can look upon them as good evidence for earlier laryngeals."

    http://www.utexas.edu/cola/centers/lrc/books/piep03.html

    ReplyDelete
    Replies
    1. I don't think Lehmann's explanation is tenable. Nobody else reconstructs *nébʰos with an initial laryngeal. Laryngeals are normally vocalised in Greek before any initial l, m or n, so the fact that there's no vowel in népʰos practically settles the matter.

      Skt. ámbhas- is semantically too distant from népʰos and difficult to connect with it formally: one can't simply postulate an isolated case of Schwebeablaut in an s-stem (*Hnébʰ-es-/*Hémbʰ-es-) and leave it at that.

      Given all these difficulties, it's more economical to assume that népʰos may make position in formulaic phrases on the analogy of other words with initial n, m or l originally preceded by *s. The resulting *hn-, *hm-, *hl- clusters were simplified not long before Homer, leaving metrical traces in set phrases, compounds and augmented or reduplicated forms. Cf. agánnipʰos 'covered with much snow' < *aga-hnikʷʰos < *...-snigʷʰ-o-.

      Delete
  14. Makes sense. What do you think of his efforts to use laryngeals to make sense of the messy developments of PIE initial *y- in Greek? They seem highly ingenious but a bit too good to be true. E.g.-

    "If we follow the procedure which Sapir used to indicate the relationship between Hitt. ḫu-u-wa-an-te-eš. ‘winds’, Gk. ἄησι ‘blows’, and αἵνω ‘I winnow’ we can arrive at a plausible explanation of the relationship between ζειά and ἤϊα. Instead of the conventional IE root wē- Sapir set up an IE base *xaweʼ - ‘to blow’. This underwent a series of ablaut modifications and changes caused by loss of phonemes which Sapir indicated as follows: (Lang. 14. 270)

    1. *xaweʼ - or *xəweʼ - : *xweʼ -
    2. *haweʼ - (or *həweʼ -) : *hweʼ -
    3. *haweʼ - (or *həweʼ -) : *ʽ weʼ -
    4. *hawē- (or *həwē-) : *ʽ wē-
    5. *awē- (or *əwē-) : *ʽ wē-
    6. *awē- : *hē-.

    For ζειά, ἤια I assume an initial γeye- : γye-. From the first of these developed ἤια; the η is a Homeric development, beside which occur also εἰαί and εἰ̑οι. From the second developed ζειά < *ζεεα, cf. ζεί-δωρος. This further evidence which we obtain from the clarification of the relation between ζειά and ἤια supports the conclusion that the root must be reconstructed with initial γ-."

    ReplyDelete
    Replies
    1. Pinault's Law says that any laryngeal (according to Andrew Byrd, only h₂ and h₃) was deleted medially between a consonant and yod. This is quite uncontroversial, especially with Byrd's correction, which accounts for the most problematic counterexamples. Byrd and Chiara Bozzone have recently argued that the same happens to any syllable-initial laryngeal before yod, also word-initially. According to them, laryngealless *j-, *h₂j- and *h₃j- all merged as *j- already in PIE, while *h₁j- survived as a cluster. In Ancient Greek, *j became /zd/ (or whatever the real pronunciation of ζ was), while *h₁j- ended up as Gk. h (via the intermediate stage of a voiceless palatal fricative). I'm inclined to accept this hypothesis (my earlier view was slightly different: I believed *h₂ patterned together with *h₁).

      Bozzone also claims that something similar happened to laryngeals before *i (so that *h₁i- > Gk. hi-), but in my opinion the evidence for this extended argument is very weak.

      Delete
    2. PS It follows that, for example, the compositionally reduced form of *h₂oju-/*h₂aju- 'lifetime, vitality' surfaced as *ju- already in PIE.

      Delete
    3. Both *h₁j- > h and *h₁i- > hi make immediate sense if *h₁ was [h]: [hj] > [ç] > reinterpretation as /h/, [hi] > [çi] > reinterpretation as /hi/.

      If *h₁ was instead [ʔ], *h₁j- > h still makes sense: [ʔj] > [ʔç] by progressive devoicing (after all, you can't voice [ʔ]) > [ç] > reinterpretation as /h/. But *h₁i- > hi does not make sense this way.

      Delete
    4. Byrd argues that /h/, as a glottal sound, is easy to coarticulate with /j/, while "Darth Vader" pharyngeals (the other two laryngeals) are incompatible with a palatal articulation in the same syllable onset, hence laryngeal deletion as an avoidance strategy.

      Delete
    5. (my earlier view was slightly different: I believed *h₂ patterned together with *h₁) [...] PS It follows that, for example, the compositionally reduced form of *h₂oju-/*h₂aju- 'lifetime, vitality' surfaced as *ju- already in PIE.

      How then do we account for /h/ rather than zeta in, say, ὑγιής?

      Delete
    6. That's one of the reasons why my earlier view was different. But there are alternative analyses of ὑγιής (with *h₁su- as the first member, as proposed already by de Saussure).

      Delete
    7. Byrd argues that /h/, as a glottal sound, is easy to coarticulate with /j/, while "Darth Vader" pharyngeals (the other two laryngeals) are incompatible with a palatal articulation in the same syllable onset, hence laryngeal deletion as an avoidance strategy.

      I think this is also true of uvulars; anyway, it makes sense.

      Delete
    8. /zd/ (or whatever the real pronunciation of ζ was)

      Apparently both [zd] and [dz]; first both at the same time, then a merger at least on the phonetic level, then perhaps a metathesis at some point... It certainly is confusing!

      Delete
  15. I'm reminded of my grandmother (a Yorkshirewoman), who pronounced words like 'human' and 'huge' with initial [j-]. Not that I'm claiming [ç] as the phonetic realisation of h₂...

    Thanks a lot for taking the time to answer all my random questions. I've decided I'm going to take some sort of online course - I'd forgotten how much I loved this stuff. So you can chalk at least one convert up to 'Language Evolution'.

    ReplyDelete
  16. It would follow that gadžo is a distant relative of Russian graždánin 'citizen'.

    Stress is on the final syllable, not the penultimate.

    It wasn't in Clackson's book, it was in Winfred Lehmann's

    I never thought much of Winnie Lehmann (as we called him at Yale), but that may be a result of interdepartmental rivalry.

    ReplyDelete
    Replies
    1. Stress is on the final syllable, not the penultimate.

      I had trouble believing this, but it's true.

      Delete
    2. Oops, Hat, you are the Master Proofreader!

      Delete