Tuesday, July 17, 2018

Schnursprecher Glockensprecher? (Kristiansen et al, 2017)

Kristiansen et al, 2017endorsed and elaborated on the hypothesis that Proto-Germanic was born in the traditional North Sea urheimat as an intermingling between the likely IE-speaking Corded Ware/Battle Axe Culture and the local, non-IE speaking Funnelbeaker farmers, among others.

This older linguistic hypothesis posits that Germanic words without clear etymologies are probably not IE and that the early sound shift in Proto-Germanic is the influence of a native tongue.  Since IE daughter languages generally have native substrates, it is reasonable that PtG borrowed as well.

While the core intent of the Kristiansen paper is focused on the genetic influence of the PtG and PIE homelands, I think identifying the Corded Ware as the prime linguistic ancestor of the people who would become proto-Germans is problematic for a host of reasons. The Germanic substrate hypothesis is vital to this, and yet it continues to be pruned back from its former status as 'an explanation for everything weird in Germanic'.

Also important to this augmented hypothesis is an outdated linguistic phylogeny stuffing Centum Germanic into a North European node with Balto-Slavic, which I'm not sure even the most bug-eyed lumper would support today.

Rather than review a host of supporting facts evenhandedly, I'll just throw darts at this narrative. If you disagree, tell me why.

Odin and Frigg (Leeke)
1. Proto-Germanic is descended from a squarely centum language, sufficiently removed from the cultural zone of Balto-Slavic and Indo-Aryan. It would become necessary to say that satemization partially baked the eastern half on the Corded Ware nation, not the western half, after CW had spread into the continent.

But if Tocharian is descended from the language of the Afanasievo Culture (Yamnaya, jr.), then it would be a remarkable coincidence that the isogloss just happens to stop at the point where lineages commonly associate with the Corded Ware nation and its descendants transitions to that of languages and lineages associated with centumization and R1b.  (Also since publishing this post, a number of early historical German tribes have been sampled and unsurprisingly, are very R1b and very Western)

2. Here's a problem regarding the Nordic Bronze Age origin that can be summarized in a single axiomatic statement. When Bell Beakers existed near other human beings, even as small minorities, the Bell Beaker cultural expression and posture is always dominant, without exception. I think this is especially true with the development of the Nordic Bronze Age and the maritime culture emphasis.

So we would have to accept that Beaker language, while present and plausibly Centum, did not permanently influence this area, but that a "Centum" Corded Ware language did.

3. But it might not matter anyway. A more refined understanding of Proto-Germanic by specialists has shown that 'germanisms' are part of a natural linguistic process over a long period of time. Some of the most identifiable characteristics of parental proto-German may be rather late.

The point is that these changes didn't happen in a decade or a hundred year period or something. They happened gradually and that that inertia continued after the breakup of proto-German and continues today, self-stimulated.

4.  Like the probable spread of Celtic in Late Bronze Age/Early Iron Age Britain, it may disappoint a lot of Germanic language speakers if the origin and spread of proto-Germanic is no older than the Jastorf Culture in which elite speakers dominated a region of archaic tongues prior to the first millennium.

This doesn't mean that Germanic is not native to the region.  Germanic certainly broke off the mother tongue in its own right and it absolutely makes sense that it dwelled in a North Sea linguistic bubble for a long time.  It may even be the case that a host of para-pre-proto-germanic (!) languages existed in the region.  

What doesn't jive is that break-up of actual proto-Germanic is just too late to envision a wide-spread "Germanic proto-nation". Somebody's axe-wielding elite dominated somebody else recently. Jastorf looks red-handed in all of this - timing, language influences, directionality.

Here's a very short chapter "The Sea and Bronze Age Transformations" by Prescott, Sand-Eriksen and Austvoll in Water and Power in Past Societies (Emily Holt, 2018)


  1. Speaking about the birth of protogermanic it is important to take note of the Schrijver's book " Language contact and the origin of the germanic languages". His theory is that proto-germanic was born when a balto-finnic population switched to a centum IE language. In doing so they pronounced this language in such a way and according to balto-finnioc pronunciation rules that a PIE centum language became proto-germanic.

    1. Schrijver's views are interesting and in a way highlight some problem areas for the above paper. I'll try and find a copy to read tonight.

      Of course it's difficult (maybe nuts) to assign language families and events to long dead people with no records. The reality could well be something unexpected like late-TRB spoke IE through elite dominance, Cw spoke some kind of Samo-Finnic and Beaker some kind of para-Vasconic. I doubt that is the case.

      Probably, TRB spoke a dead farmer language (not anything remotely related to a uralic language), CW spoke satemizing IE, and Beakers are a bit more difficult to say, but probably a centumized brand of LPIE based on its dominant daughter cultures of the regular Bronze Age, particularly in South Central Europe and early Iron Age Italy.

      While Schrijver's points on Baltic sound influences may be correct, this could have and probably did happen fairly late (possibly as the jasorf culture spread north and east) and I believe this happened after the influence of a Hallstadt Celtic superstrate, or at least a considerable degree of borrowing.

      In other words, Celtic was influencing a language that would become proto-Germanic before the first Germanic Sound shift. That would have to happen on the continental mainland with the Jastorf and the sound shift spoken of by Schrijver may have indeed happened as ppt-Germanic spread into Southern Scandinavia around 600-500 B.C.

      It would be difficult to suggest that Jastorf was the Celtic-speaking culprit because Celtic borrowings wouldn't happen in a time or place where the Germanic sound shift had already happened.

      Thanks for posting

  2. The problem with the narrative of today's historians is that they assume a linear development that actually did not exist. The tree model is as obsolete as the opposition of centum and satem languages as well as the substrate of a Pre-Germanic non-Indo-European language in Germanic. But there are influences (adstrates) from other Indo-European branches and languages (Celtic, Balto-Slavic, extinct Indo-European dialects) because of closer and more distant contacts during the centuries. So it's not as easy as we would like it.

  3. I'm nowhere near being competent to talk about this subject - though I find it fascinating anyway - but I have trouble believing that anything can be determined, given the time scale we're working with. I'm currently reading Graham Robb's book The Discovery of France, which details the incredible variation in languages and dialects that existed in France right into the 20th century. In some places, it was literally village to village, and Parisians couldn't make themselves understood in the 19th century if they left the immediate vicinity of Paris. With that kind of variety in modern, post-revolutionary France, how do you sort out what happened thousands of years ago? It seems more like a parlor game than a serious subject of study.

  4. I've got the spam filter on. Sorry for waiting to publish

  5. Late to the party - hope someone still reads this.

    When looking at the Non-IE elements in German, we first need to distinguish the following possible sources:

    1. EEF Substrate: Kristiansen e.a. have done a pretty good job in this respect. One criteria, e.g., is the "A-mobile" (e.g. German Amsel vs. Lat. merula "blackbird"), a feature both known from Hurrian/North_Caucasian, and Semitic (c.f. Span azucar "sugar").

    2. North Fennoscandian Substrate: The Saami northward expansion is rather recent, and Finnish linguists have identified a number of Pre-Saami, non-IE toponyms. Such substrate should aslo be present in Germanic, especially North Germanic. It might be worthwhile to look at North Germanic words that aren't attested in W. Germanic (and Gothic) to get a better feel about this type of substrate.

    3. Semitic super-/ adstrate: While I am anything but convinced of Vennemanns "Germania Vasconia" theory, his "Germania Semitica" has a lot going for it. Venneman, e.g., conects plough, German pflegen (to cultivate), and also folk,ON fylki "district, (army) division" to Hebr. plg "to divide", Akk. puluggu "district". Other proposed etymologies are German Adel "Nobility" (c.f. OE Æthelbald, Æthelred etc.)vs. Sem. *atsili "noble", and sib vs. Sem. *sph "family". Another point in case is the Germanic use of the Phenician "Baal" as god (e.g. Merseburg Charms) and symbol of pre-/ non-Christian origin (c.f. Baalberge ~ "pre-christian [burial] mounds"). Such adstrate would most likely go back to LBA/IA contact with Punic tin traders.
    See also https://www.caitlingreen.org/2016/12/punic-names-britain.html. Some of her examples can easily be transfered to Frisia, e.g. Rum - Roemoe - Amrum (Proto-Semitic *rwm "elevation", c.f. Ramallah).

    This still leaves numerous Germanic roots w/o clear (P)IE or Semitic etymology such as “hand”, “sing”,” blood”, “breast”, “drink”. They are not typical EEF terms, and also not reflected in Saami as one would assume if they constituted North Fennoscandian substrate. Moreover, as Uralic prohibits inter-syllabic consonant clusters, they can neither be (para-)Uralic. So, they should relate to a Pre-IE, non-EEF language spoken in the area where Germanic developed, and that was most plausibly the language of the Erteboelle Culture people. I leave it open whether Nordic Funnelbeakers maintained their Erteboelle speech (w. some EEF borrowings, of course), or whether they switched to an EEF language, into which substantial HG Substrate was incorporated. Most likely, Nordic FB spoke some sort of EEF-Erteboelle Creole.


    1. Thanks for commenting and I apologize for not seeing this in the queue for several weeks. I have a spam filter on which is driving a lot of legit comments down the list.

  6. Now to the points raised in the original post:

    A. "the origin and spread of proto-Germanic is no older than the Jastorf Culture"
    Certainly not. Proto-Germanic is defined by the Germanic sound shift (Grimm's Law), which only seems to have been completed around Augustean times. The names of the "Cimbri" and "Teutones", e.g., were not yet sound-shifted. However, the former are supposed to have originated in Himmerland (Aalborg area) on the Cimbrian peninsula, outside of the Jastorf area. As such, aside from the Jastorf area, Denmark, probably also Scania, should have been part of the proto-Germanic homeland.

    B. Some of the most identifiable characteristics of parental proto-German may be rather late.
    But others definitely weren’t. The so-called “Germanic Parent Language” (GPL) had prior to the Germanic sound shift, already undergone specific developments, a/o complete remodeling of the IE tense system (abolishing the aorist, irregular verbs of the “sing-sang-sung” type, building perfect and passive with auxiliary verbs & participle, etc.). That was quite a complex, multi-staged and time-consuming process
    Also, if Vennemann is correct in proposing Semitic ad-/ superstrate, this should have been acquired before ca. 500 BC, when the Greeks took out the Phenicians/ Punians from the Cornish tin trade. [ The onset of Punic trade with Britain is yet unclear, at least to me. It might have been as early as ca. 1200 BC, when Sweden shifted its sourcing of copper from Cyprus/ Greece to Iberia/ Sardinia.]
    These considerations take me to the Nordic Bronze Age as the latest possible point in time for GPL having begun to form.

    C. It would become necessary to say that satemization partially baked the eastern half on the Corded Ware nation, not the western half.
    But isn’t that assumption anyway made implicitly by all adherents of the “Steppe theory”, now that we know that Gimbutas’ / Anthony’s other two proposed PIE vectors, namely Baden and GAC, weren’t Steppe at all aDNA-wise ?
    Speaking of satemisation: Looking at French cent, jardin “garden”, “caille” “quail”, Modern French may be quite safely called a Satem language with a handful of centum relicts – a process that started some time after the Germanic migrations (jardin is a borrowing from Germanic) and was still ongoing when the Normans conquered Britain. Similarly, Oir has satem tendencies (e.g. cid “what”, PIE “kw”->”k” is part of the satemisation process). The centum-satem isogloss apparently isn’t very stable, to put it mildly.

  7. D.the Bell Beaker cultural expression and posture is always dominant, without exception”.
    Except, of course, in the Middle Elbe – Saale area, where BB coexisted with CWC, and the Schönfelder Group. Or Hungary. Or Italy, where BB constituted a rather short-lived phenomenon..
    BB didn’t excert dominating genetic influence in the Mediterranean, nor did they impose their language on Basques or Sardes. What you describe applies mainly to Britain.

    E. So we would have to accept that Beaker language, while present and plausibly Centum, did not permanently influence this area.
    Hmm.. BB was a multicultural, genetically and probably also linguistically diverse network. We may, however, rightly speak of a “Western BB language”, given the extent to which British Neolithic DNA was replaced by incoming BB.-That “Western BB language” seems to have formed somewhere in the Dutch-German plain between Oostwoud, Noord-Holland (Olalde e.a. study), and Quedlinburg (the Irish Rathlin study). If that language was IE (which seems likely), it arose in the Single Grave Culture from interaction of incoming Steppe populations with the TRB West Group and/or the Bernburg Culture.
    Now, If we follow Kristiansen, GPL emerged from interaction of the Danish Single Grave Culture with the TRB North Group. How different, do you think, both these languages should have been, considering that TRB West originated from TRB North expansion (see https://adnaera.com/2018/09/09/a-first-and-intriguing-glimpse-at-trb-west-group )?
    IOW – I deem it quite plausible that BB West, as BB North, spoke some kind of Proto-GPL. Cultural contact across the North Sea continued during the BA, as e.g. signified by identical halberd forms in Britain and Denmark, and Unetice’s economic base was trade of Cornish tin towards the Danubian basin and ultimately Greece/ Crete/ Egypt. Consequently, we may have had several para-PGL languages spoken by the LBA in Britain, N. Germany and Denmark, but after a good millennium of differentiation, they should still have been closer to each other than English, Dutch and Danish are today.

    F”Like the probable spread of Celtic in Late Bronze Age/Early Iron Age Britain ..”
    This essentially nails it, albeit in an incomplete form. There were two expansions, one related to Hallstatt, the other one to La Tene. The Hallstatt expansion brought non-Celtic ethnonyms such as “Picts”, “Parisi” (Yorkshire) and “Venicones” (Firth of Tay) to Britain; the latter can be plausibly connected to East Alpine ethnonyms such as Veneti, Vennones, or Vindelici. It also encroached Central Europe (Pomeranian Face Urn Culture, House Urn Culture). The former ultimately migrated to the Vistula to form the Vistula Veneti, the latter is responsible for certain Italic toponyms in Thuringia, e.g. Jena (from PIE *ǵónu “knee, (river) bend”, cf. Genua with the same etymology). The disputed “Nordwestblock”, identified from non-Germanic / non-Celtic toponyms, may also belong here, as do Iberian non-Celtic-IE languages such as Lusitanian.
    The Celtic expansion can be equaled to the archeologically well attested La Tene Culture, and reached at least North England (“Brigantes”), as the southern end of the North German plain (and of course, Celtiberia, Bohemia, the Cracow area, Galatia and a number more places).
    However large or small the GPL, “maritime beakers” area once may have been, the combination of both IA expansions essentially reduced it to Jastorf plus S. Scandinavia. Speculatively, a few para-PGL isles may have remained. The Irish Cauci around Dublin, e.g., have often, e.g. by Pokorny, been connected to the North German Chauci. However, while the latter were already undergoing the Germanic sound shift when noted by Roman authors, the former ethnonym appears to have remained unaffected (etymologically, both names can be connected to PGerm *hauhaz [GPL *kaukaz] “high”, w/o further IE parallels).

    1. "BB didn’t excert dominating genetic influence in the Mediterranean, nor did they impose their language on Basques or Sardes. What you describe applies mainly to Britain."

      By a dominant posture I mean that their presence was apparently tolerated in the lands of others, especially with in regards to access of resources, farmland and avenues. I favor an intrusive interpretation, but other scenarios are possible.