PIE *bherg- and *bhergo

Discussion in 'Etymology, History of languages and Linguistics (EHL)' started by terredepomme, Apr 21, 2012.

  1. terredepomme Senior Member

    Human Language
    Are *bherg-(height, cf. Eng. berg) and *bhergo(protect, cf. Rus. беречь) any related?
     
  2. Ben Jamin Senior Member

    Norway
    Polish
    I don’t think English ‘berg’ and Russian ‘беречь’ are related, but Scandinavian “berge” (to save) is ertainly related to ‘беречь’.
    ‘Berg’ is related to Russian ‘bereg’ (rand, riverbank, seashore).
     
  3. berndf Moderator

    Geneva
    German (Germany)
    On what grounds?
     
  4. Ben Jamin Senior Member

    Norway
    Polish
    Well, it was a too hasty statement. I knew the relation between “berge” and‘беречь’, and between "berg" ened "берег", but I didn't know the relation between "berge" and "berg". I found the explanation however in the Grimms' dictionary: "bergan" meant to save something to the shore, hence it is related. The relation, however, is not direct: ‘беречь’ is related to "berge/bergan", and the latter is related to "berg".
     
  5. berndf Moderator

    Geneva
    German (Germany)
    I am not sure; that's why I asked. The meanings of the verb stem -berg- in German range from recover via rescue and protect to conceal. All of these meaning have an obvious connection to беречь. From the attested history of Germanic languages I can't see a way to decide, if the verb stem -berg- and the noun stem berg- (=mountain) are cognate of just the result of a phonological merger. Germanic /g/ can be derived from PIE /g'h/ or from PIE /gh/.
     
  6. Ben Jamin Senior Member

    Norway
    Polish
    You don't buy the Grimms' explanation?
     
  7. berndf Moderator

    Geneva
    German (Germany)
    He argues with Adelung whether bergen is derived from Berg (Adelung) or Berg from bergen (Grimm). But neither of them discuss the possibility that the two might be unrelated. I find this not entirely satisfactory.
     
  8. Dhira Simha Senior Member

    UK
    Russian
    I also believe that Rus. bereg "bank, shore" and beregti "to preserve" are related. This is how this word appears in the Russian-Sanskrit Dictionary of Common and Cognate Words


    52 берег byeryeh byerega bhṛ भृ
    bereg bank, shore to bear, carry, convey, hold
    Traditionally, this word is linked to the SA bṛhat बृहत् ‘large, lofty, high, tall, great, large, wide’ and the Avestan berezô (berez) 'high, great, lofty' (Vasmer, I, 153) on the basis of its supposed relation to the GER berg ‘mountain’. Semantically, such parallel is rather dubious. On the other hand, there is a more obvious similarity with the RU беречь bereč ‘to protect, guard’ so the actual SA cognate root scold be bhṛ भृ meaning ‘to bear, contain, possess, have, keep; to carry off or along’. One can imagine river banks as containing the waters and carrying them along. They could also be viewed as protecting people from water, which was often considered deadly (see море more). If one looks at the etymology of берег bereg from this angle - its connection with the verb беречь bereč ‘to bear, contain, possess, have, keep; to carry off or along; to protect, guard, take care of’ becomes logical. Cp. also the dialectal RU берёга berёga 'care, protection'. See беречь bereč. UA бе́рег; BG брегъ; SRB бриjег; SLO brėg; CZ břeh; SK breh; PL brzeg; U.LS brjóh; L.LS brjog. N 3
     
  9. Ben Jamin Senior Member

    Norway
    Polish
    How strange that the connection between the two words should develop independently both in Russian and Germanic.
     
  10. aruniyan Senior Member

    Tamil
    Remove this post if its some sort of novel theory, but like to bring in three Tamil root words that are interesting.

    peRu, peRiya, peRuku - Big, Large, Increase etc...

    peru, Perra, perrOr - Receive,Possess, Parents(those who receives)

    pira - Birth


    peRu(large)-noun and peru(Receive,Possess)-Verb.
     
  11. Dhira Simha Senior Member

    UK
    Russian
    These words appear to relate to the verb peRu-tal "to get, obtain, secure, possess; to bring forth, bear as children; to be get, generate" so the meanings "peRu, peRiya, peRuku - Big, Large, Increase etc " are probably secondary following the logic: to bring forth, bear > grow > increase > became large > large. It is is an interesting fact but Tamil being non-Indo-Aryan any similarity should be treated with caution. A good example is the Skr. moraka "the milk of a cow seven days after calving" which sounds exactly like Rus. moloko "milk" (r/l interchange is common on Skr.) but, in fact, it has nothing to do with it. It is off-topic, but as an exercise, try to explain why. If you are proficient in Tamil you will see the reason.
     
  12. arsham Senior Member

    Canada
    Persian
    Classical Persian has borz meaning height from the same root as boland meaning high. Middle Persian has the superrelative bâlist meaning highest in addition to the aforesaid forms!
     
  13. Dhira Simha Senior Member

    UK
    Russian
    Sorry, I do not quite see to which of the earlier posts it refers. Are you sure these three Persian forms relate to the same root? My first impulse would be to associate "borz meaning height" with Skr. varṣiman 'height, top' and Rus. вершина veršina having the same meaning. Both relate to the root vṛh वृह् (more often having the form of bṛh बृह्) 'increase, expand, further' that is directly cognate with vṛdh वृध् ‘to grow up, to rise, to elevate’ and the RU верх verx ‘top’. The other two words, you quoted may be related to Skr. bala "power, strength, might". In the cognate Rus. boljš the meaning has shifted more towards "big" and, in some contexts, "tall". It appears directly related to bâlist and both words are connected with the ancient superlative suffix -iṣṭha. I doubt that any of these have relation to PIE *bherg- and *bhergo.
     
  14. fdb Senior Member

    Cambridge, UK
    French (France)
    These words belong to *bherg-. –rg- becomes –rh- in Sanskrit, –rz- in proto-Iranian, -rd- in Old Persian, -l- in Middle and New Persian. Thus we have Avestan barǝz- in ablaut with bǝrǝz- ‘high’, barǝzah- ‘mountain’, Persian buland ‘high’, bālā ‘height’, bālist ‘highest’ etc. burz ‘high’ is a non-Persian (Parthian) loanword in Persian.
     
    Last edited: Jun 23, 2012
  15. arsham Senior Member

    Canada
    Persian
    I second this thorough explanation. The Middle Persian form of bâlâ is bâlây and burz in shahnâmeh is always used as a noun!
     
  16. Dhira Simha Senior Member

    UK
    Russian
    As I said, I am not an expert in Persian so I may accept your explanation of bālā ‘height’. However, there are some questions: what would be the Slavonic reflex of -rg- ? How does Skr. varṣiman 'height, top' and Rus. вершина veršina fit here and how do they relate to bereg "bank, shore" and beregti "to take care of, to spare"? Also the interchange x - š is common in Slavonic but I never met it in Sanskrit. Skr. /h/ is normally preserved before /i/. Why did it have to change to -ṣ- in varṣiman and even in varṣman? If we apply RUKI law (*s > (ʃ) > *x.) then we should consider the original form * brs/vrs because we know of now examples of /h/ going to /s/ in Sanskrit. How does this tie up with your presumed "–rg- become –rh- in Sanskrit"? The words varṣiman and varṣman are the only link between the notions of "height" and the cardinal meaning of bṛh "to grow thick, increase". I can see the connection varṣiman - barǝzah - veršina but how does the hypothetical *bherg fit in here? I would really like to understand this.
     
  17. fdb Senior Member

    Cambridge, UK
    French (France)
    Iranian *bṛz- : *barz , Avestan bǝrǝz- : barǝz- are cognate with Skt. bṛhat- , barh-aya-, from IE. *bherg-, with bh- > b- according to Grassmann’s law.
     
    Last edited: Jun 23, 2012
  18. Dhira Simha Senior Member

    UK
    Russian
    Does this effectively mean that the true reflex of IE -rg in Skr. is -rs and not -rh? With RUKI law it would become -rṣ but how does the /h/ in bṛh fits here? The change of š to h (x) is specific to Slavonic, as far as I know.
     
  19. fdb Senior Member

    Cambridge, UK
    French (France)
    No, it does not mean this at all. The Skt. words with ṣ do not belong to this root.
     
  20. Dhira Simha Senior Member

    UK
    Russian
    It looks as an ad hoc solution to me. Avestan barǝzah, barǝziman and barǝzišta are considered relating to bh but Skr. variman which exactly coincides in form and meaning with barǝziman should not be related because it just does not fit? Interestingly, a form varhiman 'height, top, surface, uttermost part' was attested in RV. as well and it is believed to be a duplicate of variman. Now, varh- here looks suspiciously close to vrh/bṛh. In fact, the guṇa form of vrh would be exacly varh. I do not think you have a ready answer, I just wanted to say that things could be more complicated that it appears.
     
  21. berndf Moderator

    Geneva
    German (Germany)
    Coming back to the original question, we have the Germanic stem berg- which according to Grimm's law should have the PIE etymon *bherg- plus we find probably Slavic cognates (discussed above) and now we are presented with an Iranian root with the same presumptive etymon. If the other Russian and Sanskrit words you threw into the discussion belong to the same group or not might be interesting but without further proof that the entire group should be derived from a completely different root the is not essential to the question here.

    There are two semantic ranges for this group: Hill/mountain/top/high on the one hand and protect/cover/bring ashore/shore on the other hand. The question that was asked was if these ranges originate from a single meaning and a single etymon or if those are distinct roots which just happened to have been become indistinguishable.
     
  22. Dhira Simha Senior Member

    UK
    Russian
    The additional evidence is essential if we want to resolve this. On one hand, we have a group of supposed cognates which are conveniently listed here *bhergh and which appear to cover most of the IE groups except Balto-Slavonic. The Slavonic *bergъ "bank, shore" is controversial and there may be at least one intelligible way of explaining it as deriving from a different root than *bhergh 'mountain?'. On the other hand, we have a very similar beregti 'to preserve, to take care of', which may also be reconstructed in a similar form *bhergh(?) with the general meaning "to preserve, protect, nourish, care for". Semantically, these meanings are hardly compatible with "hight, top, mountain". Also this root is confined to only some Slavonic and some German dialects (Goth baírgan, Germ. bergen). The Eastern Lithuanian bir̃ginti "to save, protect" may be a Slavonic loan. Such limitation is suspicious and needs explanation.

    If we look into the Old Indo-Iranian, particularly Vedic we find there an exact semantic and a near exact phonetic match: bhṛ - bhárati which has two principal meanings "to bear, carry, convey, hold" and "to support, maintain, cherish, foster". The latter is fully compatible in meaning with Slavonic bereg- and German bergen. The difference is only in one final g(h). Compare Goth. baírgan 'to conseal, hide' and baíra 'I carry'. Clearly we deal with two closely related roots and all we need to do is to explain the final g/h in Slavonic-Germanic. An interesting clue may be the Rus. infinitive bereč 'to preserve, take care of' where č = palatalised ending t(i) and the -g is missing (cp. beregu 'I preserve, take care') vs. a parallel form beregti.

    We have, therefore, to posit two separate closely sounding roots. In fact, in Pokorny they were shown as distinct roots *bhereg'h 'high; mountain' and *bherg'h 'hide, keep' but other dictionaries give them as one form *bhergh (American Heritage Dict. of IE Roots p.11). I do not know where the *bherg and *bhergo which are in the title of the thread come from. Next comes the interesting discussion on the extremely important Vedic root bṛh (which is the cornerstone of the Brahman concept) and whether or not we can include it here etc. but this is off-topic.

    Conclusion: *bhereg'h 'high; mountain' and *bherg'h 'hide, keep' are two independent roots. The only possible link to connect them could be the Slavonic bereg, breg 'bank, shore' but its etymology is ambiguous and controversial.
     
    Last edited: Jun 24, 2012
  23. berndf Moderator

    Geneva
    German (Germany)
    I am not yet convinced. There is some semantic overlay which needs to be explained, like Germ. Burg/Engl. borough which is supposed to be derived from the meaning mountain/hill but which could equally be derived from the meaning to protect.
     
  24. Dhira Simha Senior Member

    UK
    Russian
    O.E. burg, burh "a dwelling or dwellings within a fortified enclosure," from P.Gmc. *burgs "hill fort, fortress" (cf. O.Fris. burg "castle," O.N. borg "wall, castle," O.H.G. burg, buruc "fortified place, citadel," Ger. Burg "castle," Goth. baurgs "city" - primarily means "enclosed place" not "hill". Not to mention that there is a big semantic difference between Hügel and Berg. Any connection with Berg "mountain" is probably a later development as such enclosed forts were usually made on elevated places. I am not a specialist on Germanic etymology but I immediately see a direct link with Lat. muro and Skr. mura 'encompassing , surrounding'. Please explain why in all these cases there happened a change of /e/ to /u/ especially if the word Berg was preserved perfectly? Voicing or devoicing of initial labial is a trivial thing. Often it is purely psychological and a matter of spelling. Another possible connection would be Skr. pur 'a rampart, wall, stronghold, fortress, castle, city, town' preserved perfectly in Slavonic -por: zapor, upor etc. (and probably Eng propp and Germ. pfropfen). I would not even exclude that Skr. mur and pur could be phonetic variants.
     
  25. eamp Junior Member

    Vienna
    German (Austria)
    LIV (Lexikon der Indogermanischen Verben) distinguishes two verbal roots here *bherĝh meaning "become high, rise" which is preserved in verbs with this meaning in Hittite, Tocharian and Armenian apparently. Many languages have derivations from this root meaning "high" or "hill".
    And on the other hand *bhergh, with non-palatal gh, with meaning "preserve, care for", but it's listed with a question mark. Besides the obvious Germanic and Slavic words (and not so obvious like English "bury" and "borrow") it also connects an iranic word meaning "to honor".

    If this construction is correct Slavic "bergъ" can't be from the root meaning "high" though, we would expect "berzъ" unless it is a loan from a western (non-satem) dialect.

    For Germanic *burg- I think derivation from *bhrĝh is most likely since the formation (zero grade root noun) seems to have an exact parallel in Celtic *brigs "hill", maybe also in Iranian, but I don't have a dictionary for that. It would then later have been influenced in meaning by the root meaning "protect" as well as the Latin "burgus".
     
  26. eamp Junior Member

    Vienna
    German (Austria)
    The chance of /e/ to /u/ is because of a well known phenomenon in Indo-European languages called ablaut. In this case it represents an earlier difference of *bherĝh and *bhrĝh.
    Latin "murus" (which is what you mean I guess) is attested in earlier forms "moerus" so it's from *moiros, which the Skt. word definitely can't be derived from. Your other words also have nothing to do with each other.
    Such willy-nilly word comparisons as you offer have not been accepted in Comparative Linguistics for at least 150 years, sound changes in general follow regular patterns and are not just speakers substituting one sound for another randomly.
     
  27. Dhira Simha Senior Member

    UK
    Russian
    I remember from my school days that Lat. burgus is actually a late Latin borrowing from Germanic. Please correct me if I am wrong. How could it influence Berg to turn it into Burg? Again, the principal meaning of burg is an enclosed fortress and not "hill". Sorry, I can understand bury and borrow but what does 'honour' have to do here? Is it not suspicious that so many various words seem to derive from the same mythical *bhergh?
     
  28. berndf Moderator

    Geneva
    German (Germany)
    Don't get me wrong. I don't say, you are wrong. I only think that we don't know yet. Grimm and Adlung (as mentioned above), e.g. both assume the two roots to be related but disagree about which one is primary. As to Burg, the common meaning it fortified castle/settlement on a hill. Here it is given as derived from high.
     
  29. Dhira Simha Senior Member

    UK
    Russian
    It is just another unprovable theory. If burg is a derivative of the same mythical *bhergh please explain why this root manifests itself as burg and berg in the same language group?
     
  30. berndf Moderator

    Geneva
    German (Germany)
    Just like yours, and this was my point.
    Well, here you have the testimony of a native speaker because in modern German Burg still has this meaning (castle on a hill). And multiple forms derived from the same root aren't so rare. E.g. because they ware derived from different dialects, e.g. English chase (from French) and catch (from Norman French).
     
  31. perevoditel Junior Member

    I would only say that it doesn't fit to Slavic languages, because "mountain" is "gora/góra" in most of them. "Versina" means "top", there are also derived terms in Polish to that.

    And what about "bjerk"/"björk" meaning :"birch" (kind of tree)? Is it also derived from PIE?
     
  32. Dhira Simha Senior Member

    UK
    Russian
    You mean Burgberg? And Bürger is a a citizen of a "castle on a hill" or just somebody living within city walls?
     
  33. berndf Moderator

    Geneva
    German (Germany)
    A Burgberg is the hill on which a Burg has been built. The meaning Burg=city does not exist in German and isn't attested in any development stage. Originally, a Bürger accordingly was someone who lived in a castle. The extension to inhabitants of a city is conventionally explained by the fact that German cities were usually built around castles. You find this explanation e.g. in Grimm. Later, the term was, probably under French influence extended further to mean citizen of a country.
     
  34. Dhira Simha Senior Member

    UK
    Russian
    Having pondered on the issue I should admit that there indeed could be a deep connection between the two meanings "mountain, hight place" and "to preserve, protect". Currently, there is no agreement on *bhergh and, as I can see from Trubachev's dictionary of Slavonic Inherited Lexicon, he wisely distanced from expressing his own opinion. The general consensus is that both German Berg and bergen and Slavonic b(e)reg, beregti are ancient inherited words. I can also see your point on Burg. Still the Skr. bṛh in which the cardinal meaning is "to expand, grow" and "high" being only a marginal meaning needs further study. Due to the potential instability of the initial labial (cp vṛh/bṛh) I would suggest to view Skr. vṛjana "an enclosure , cleared or fenced or fortified place (esp. `" sacrificial enclosure "' ; but also `" pasture or camping ground , settlement " inherently connected with the verb vṛ 'to cover , screen , veil , conceal , hide , surround; to ward off , check , keep back , prevent , hinder , restrain' which is, by the way, believed to be related to varṣiman discussed earlier. At least this opens some alternative path which has not been explored yet, to the best of my knowledge. One more clue could be the existance of parallel forms bereginja and vereginja 'benevolent female spirit-protector' in Russian (see ESSja 1-193).
     
    Last edited: Jun 25, 2012
  35. berndf Moderator

    Geneva
    German (Germany)
    It seems we are on the same wavelength now.
     
  36. Dhira Simha Senior Member

    UK
    Russian
    I think that the "classic" idea of "reconstructing" PIE roots as CVC radicals, modelled from Semitic may actually be short-sighted. IE languages, particularly Skr. are known for extensive use of compounding. Can we imagine it stretching deeper into the root structure? I do not mean pre-fixation, this is what is usually meant by compound verbs. Looking at Sanskrit we can see numerous combinations of simple V of CV roots with such radicals as da "giving", dhe "having, placing, ga "movement", ka "comparison" etc. For example we have an element va which has a variety of meanings ( which partly can be explained because it may be a fusion of va, vo and ve) meaning, particularly 'air , wind ; addressing ; reverence'. Then we have a number of more complex formations:vac(k) - to speak; vad - to speak (va + da?); vaT 'to speak' ; vich 'to speak' Similarly, with vRj we can see a fusion of vR and j. Perhaps by endlessly creating various similar CVC roots we are chasing fathoms? It is off-topic, I am just reflecting.
     
    Last edited: Jun 25, 2012
  37. eamp Junior Member

    Vienna
    German (Austria)
    That Burg never meant city isn't true though, in Old High German in fact this was the normal meaning. In glosses it appear for "civitas" and "urbs" also for Greek "polis" and in the Bible translations the cities are all called "burg", Rome is called a "burg" too. Of course all those will have been fortified cities with a wall made of stone and most were originally built on hills too. It could also translate "castellum" and "castrum" apparently, but those occur more rarely.
    "burgari" and "burgliut" with the meaning inhabitants of a city also dates from Old High German and this actually predates the development of castles and then cities around castles in Germany. So I would say the meaning "city" (maybe more originally ~fortified settlement?) is definitely old in German it just became restricted to the modern meaning of castle sometimes in the middle ages when it was replaced by "Stadt" in the meaning of city (ohg. "stat" still meant just generic "place").
    Interestingly, when I researched I found that in Old Icelandic "borg" still has three meanings: "castle, fortification" and "city" but also "hill".
     
  38. berndf Moderator

    Geneva
    German (Germany)
    Yes, I think so.
    You are right. My mistake. It was replaced in that meaning by Stadt only in MHG.
     
  39. Dhira Simha Senior Member

    UK
    Russian
    I am afraid, we are stuck at this point and will only go in circles. I have nothing to add. This discussion was a good experience for me, though. I learned a lot. Thanks to everybody!
     
  40. berndf Moderator

    Geneva
    German (Germany)
    I am afraid so.
    So have I. Thanks a lot.
     
  41. aeneas dardanus Banned

    Ulpiana
    Dardanian
    You are right;
    The previous is the original sema of the root in discussion;
    whereas the later is the derived meaning of the first, -therefore a late (derivative) loan on eastern Slavic.
     
  42. berndf Moderator

    Geneva
    German (Germany)
    This has been the point of controversy in this thread. Just claiming one side is right and the others are wrong won't do the trick. Please substantiate your claim.
     
  43. Dhira Simha Senior Member

    UK
    Russian
    I fully support berndf. I can clearly see his point and keep on thinking on this. There are too many verbs with similar phonetic structure in Eastern Slavonic:
    стеречь - стерегу; бечь - бегу; жечь - жгу etc. so беречь - берегу does not stand out in any way to be considered a loan. At this stage I am still inclined to separate the two meanings and posit separate roots but this is my personal opinion as of August the 7th, 2012 at 23.55 UK time :)
     
  44. aeneas dardanus Banned

    Ulpiana
    Dardanian
    (controversy; I know there is none, at least, not among scholars).

    I think that the meaning of the root {b:rgᵏ < > ᵖbr:gᵏ } is very well augmented and explained in signifying a (round hill); hillock; mound(1); bank…; -either in a flat field, or beside water where troubled communities found it a better place to make a stand, and in time: fortify(2), and later on permanently settle there for: safety(3) reasons.

    I presume that there is no point in providing that: 3. conceptual notions, and 2. synonyms, trigger their names from actual 1. objects, - not the other way around.


     
  45. berndf Moderator

    Geneva
    German (Germany)
    Grimm and Adelung (see #7 above) agreed that they are related but had opposing views which meaning is derived from which. A good century later, Pokorny lists them as two different root (here, #239 and #245). If you think that new evidence has decided the matter since their days, please share your information with us.
     
  46. Treaty Senior Member

    Australia
    Persian
    I'm not sure about the Russian bereg and its possible Slavic cognates or how words change in Russian. Nevertheless, the first thing that it reminded me as "bank" or "shore" was Persian bar (= side; for water: bank, shore) and its Greco-Latin cognate para. Anyway, it doesn't explain the end "g".
     
  47. Ljudevit New Member

    Croatian - Croatia
    In Croatian "brijeg" means hill, {"breg" in kajkavian Croatian). There is also "brdo", ("berdo"-kajkavian) meaning hill, but that might be a stretch. I believe that Bulgarian has "bair", "баир".
     

Share This Page