14 Mar 2008

Proto-Semitic as a second language

This will be something to ponder for the weekend. It's another crazy idea I had that may not be so crazy. Let me just first spit out the revelation I'm having and I'll explain it all afterwards.

I will start with the claim that Proto-Semitic originated from the Syria-Palestine area[1], rather than from Southern Arabia as has been so often claimed. Then, considering the well-known fact that Neolithic innovations originated from Western Asia and only later spread into Europe, I'm going to suggest that Proto-Semitic speakers were not only people with agricultural know-how, but that their language became a vibrant trading language well beyond their immediate area. What I'm suggesting is that multilingualism was not only common during the Neolithic but even vital for communities and their material well-being. I don't know why I didn't clue in before, but if Proto-Semitic speakers were ahead of everyone in terms of technology, naturally their language too might become a hot commodity. And if knowing that language was in demand for trade, then it follows that there were large areas surrounding the immediate Proto-Semitic language area where people would have adopted Proto-Semitic as a second language!

Think about it now. Around 5500 BCE, speakers of "Mid Indo-European" (MIE), ancestral to later Proto-Indo-European (PIE), might have been situated further into the Balkans to take advantage of goods coming in from the south, perhaps along the coastline, and these people would have been at least semi-fluent in Proto-Semitic in order to communicate with the incoming traders. (I mean, how else could they likely communicate with each other other than becoming bilingual?) The Syria-Palestine area was afterall a center for agriculture and we know that there are words in Proto-Semitic relating to agriculture as the American Heritage Dictionary explains in detail: "There are many Proto-Semitic terms referring to agriculture, which was a significant source of livelihood. Words for basic farming activities are well represented: fields (*ḥaql-) were plowed (*ḥrθ), sown (*ðrʕ), and reaped (*ʕƛd); grain was trampled or threshed (*dyš) and winnowed (*ðrw) on a threshing floor (*gurn-), and ground (*t’ḥn) into flour (*qamḥ-)." As for multilingualism in ancient communities, this shouldn't be much of a shocker considering the examples of Quechua and Swahili. Multilingualism was much more common in ancient times than we often appreciate.

My idea hopefully will raise mindteasing issues concerning the loanwords in PIE. Are they really the product of direct Proto-Semitic contact or is it even more direct than I thought. That is, are these loanwords rather the natural result of generations of bilingual speakers of both languages? Bilingual interference like this happens all the time and I can speak with authority on that, being bilingual, that it is a common tendency for two languages that reside in your brain to jumble together sometimes, producing spontaneous loanwords. I remember the time I accidentally blurted out the word *distach instead of detach while talking to a friend. I realize now that this was because of a subconscious mental association between the French prefix dé(s)- and English de- or dis-. Even though my native language is English, French obviously had an affect on me. It only took a second for my tongue to say it, but hours to reflect on the implications of my dysphasic faux pas. Unlike in the modern age of grammatical simulacra, there would be no anal-retentive grammarian in the Neolithic to stop a bilingual person whose mother language was PIE from slipping up now and then, using a Semitic word for a PIE word that he or she may have momentarily forgot. It would only be a matter of time for the products of this language switching to become normalized in a community and spread to neighbouring communities.

I have tonnes more ideas on this, such as the "borrowing hierarchy" issues in relation to intensive language contact which were brought up in Elsik and Matras' Markedness and Language Change: The Romani Sample (2006) which I cited in an earlier post, but I'll just have to take this one blogpost at a time. Breathe in, breathe out, breathe in... Hehe.

NOTES
[1] Lipinski, Semitic Languages: Outline of a Comparative Grammar (2001), p.43 (see link): "Since the Semitic languages go apparently back to a common origin, the question of the location of the speakers of this Proto-Semitic language has been often considerederd of importance. Various regions have been taken into account: Syria, Arabia, and Africa." Sadly, Lipinski seems to overrely on geographical names to draw a conclusion about the likelihood of Semitic speakers in Syria. He seems here to be ignoring or is ignorant of the important issue of Semitic loanwords in PIE altogether.

UPDATES
(March 14 2008) I should just clarify something just in case people incorrectly add two plus two together to make five. The graphic above is not meant to endorse in any way the "Out-of-Anatolia" fringe hypothesis promoted by Gamkrelidze and Ivanov, which is in my view assuredly wrong. I've been convinced by Alan Bomhard's view that Indo-European originally came off the steppelands of Western Asia from the east. However, I'm tossing around a new idea that Late IE was largely an offshoot of northern Mid IE dialects (although I haven't the foggiest idea how I could prove this!).

10 comments:

  1. Quite convincing all this.

    Though I wonder if there's no such thing as linguistic purism in the Neolithic age. People may have had a less clear view on what is 'grammatical' and what is 'pure', but linguistic purism to me seems of all ages and times; no reason for it not to exist in those times.

    The big problem with that of course is: Only proof we have of linguistic purism is ever since people started writing; so then of course it's hard to say if purism develops with an orthography (and thus also a better grasp of grammar/language)or whether it's just something of all ages.

    ReplyDelete
  2. Phoenix: "Though I wonder if there's no such thing as linguistic purism in the Neolithic age."

    Things that make you go hmmm :) However, when I was talking casually about modern purism, I was thinking more of examples like the Quebec "Language Police" (which shames us as Canadians, alas) or the more covert avoidance of foreign words by Finnish speakers (e.g. "telephone" -> puhelin). These instances seem to be all products of modernday nationalism and ethnocentricity, things that didn't and couldn't exist in the Neolithic.

    At best, I'd imagine a kind of general xenophobia (i.e. foreigners versus one's band/tribe), but since we're exploring Semitic loanwords and the impact this language had on PIE, naturally there wasn't enough purism to speak of, was there? ;) The very reason Pre-Indo-European speakers acquired a pastoral lifestyle and let go of their former hunter-gatherer ways was precisely because of their open trade with others overall, not their xenophobia.

    As for grammatical purism, I sincerely doubt that IE speakers even had words for "noun", "verb" and "adjective" to make much of a fuss. At most, perhaps there could be neighbouring rival bands or tribes that had certain idiosyncratic speech patterns or vocabulary (i.e. shibboleths, if you will) that a group might feel should be avoided but I can't think of anything more involved than that. But maybe I'm not thinking straight because it's morning and I have no coffee in my hand.

    Could there be other forms of language purism that might have conceivably existed in the Neolithic other than what I suggested?

    ReplyDelete
  3. Hello, I'm a blogger who has a big interest in the Neolithic expansion and North Africa, and I'm currently looking into the languages involved in the population migrations..

    It seems to me more likely that Southern Anatolia/Northern Iran and not Syria would be the starting point of an expansion for proto Semitic, as this is the point of origin for the early Neolithic crops and population expansion - the near east is NOT where farming originated or where the first farmers expanded from.

    I'm modelling this on the expansion of the Bantu languages with farming. It might explain why Celtic langages have some similarities to Afro Asiatic grammar; they could have retained aspects of the grammar when the IE languages arrived.

    There was also an expansion from Anatolia about 13,500 years ago, so some of the older Greek and Med Island languages like Minoan could have been cousins to Semitic, as they were settled directly from Anatolia.

    Afro Asiatic was probably imported from North east Africa by a population movement that arrived in Southern Anatolia about 16k ago - the expansion of the Halfan culture, which had an offspring culture in Syria called the Natufians, which parented the Belbasi culture in Southern Anatolia, which in turn colonised Southern Greece and Crete.

    Just thought I'd add a little info about ancient population movements into the debate. I think PS expanded from Anatolia about 9k ago, and is probably related to a lot of the pre IE languages aorund the Med, as a parent or cousin.

    Your blog has been very interesting. If you ever want any info on population movements in the area of north Africa and Anatolia that might be relevant to Afro Asiatic, then just ask, as it's pretty much my specialist area and they do shed some light on the possible paths taken by ancient languages.
    http://mathildasanthropologyblog.wordpress.com/

    ReplyDelete
  4. Hi, Mathilda. Thanks for your thoughts. And now the battle begins <;-)

    mathilda: "It seems to me more likely that Southern Anatolia/Northern Iran and not Syria would be the starting point of an expansion for proto Semitic [...] the near east is NOT where farming originated or where the first farmers expanded from."

    Don't confuse the ultimate origins of agriculture with the location of later agricultural Proto-Semitic speakers themselves. Different subject, different time period. Syria is a reasonable means of linguistic transmission via Anatolia and doesn't conflict with basic archaeological evidence that shows that the Balkans adopted agriculture through Anatolia (see Bellwood, First Farmers: The Origins of Agricultural Societies (2005), p.74). So I'd estimate that this language contact must have occurred somewhere between 6500 and 5500 BCE (ie. long after agriculture had already started diffusing outward from its source). Note the Proto-Sesklo culture.

    mathilda: "There was also an expansion from Anatolia about 13,500 years ago, so some of the older Greek and Med Island languages like Minoan could have been cousins to Semitic, as they were settled directly from Anatolia."

    The "Minoan-is-a-Semitic-language" hypothesis is, as far as I'm concerned, in the same pseudoscience category as "Intelligent Design". There is nothing to show that Minoan is Semitic whatsoever and the people who have tried (eg. Cyrus Gordon) have failed miserably with unacceptable methods and a heap of incredible assumptions. The general consensus among academics is that it's related to neither Indo-European nor Semitic languages. I suggest that Minoan is related to Etruscan here on my blog but, to be fair, my view cannot be proven yet either.

    mathilda: "Just thought I'd add a little info about ancient population movements into the debate. I think PS expanded from Anatolia about 9k ago"

    Yes, thanks for the info, but that date is far too early for Proto-Semitic. A date of around 5500 BCE seems to me to account better for what would otherwise be a suspiciously slow rate of language change between the date of the protolanguage and the first record of Akkadian.

    ReplyDelete
  5. The signs point a lot more towards an Arabian origin (keeping in mind Syria is the northern-most tip of the Arabian peninsula anyway).

    The further south one goes into the peninsula, the more pristine the Semitic features of the languages there become. In fact the only Semitic languages still retaining all 29 proto-Semitic phonemes exist in Southern Arabia (MSA languages), and modern Arabic itself retains 28 of these. Compare that to almost all of the northern Semitic languages, and we find they lost dramatic numbers of phonemes, indicating the language spread north and upon contact with other languages (perhaps IE languages) the distinct Semitic phonemes quickly eroded.

    Of course Ugaritic bucks this trend, but there's ways to explain that.

    ReplyDelete
  6. Abu: "[...] and we find they lost dramatic numbers of phonemes, indicating the language spread north and upon contact with other languages"

    Nonsense. Proximity to original locale has nothing to do with rate of phonemic change or any other change and I won't indulge you in that invalid argument. If your view were valid, clearly Armenian should be closer to PIE than Tocharian which migrated far into Asia. If it were valid, how could Quebecois retain archaicisms of 16th-century French? Rate of change obviously has many factors, both external and internal.

    An important fact I use to strong-arm the issue is that if Proto-IE *septm 'seven', an internally unanalysable numeral, is an accepted item of the oldest lexicon circa 4000 BCE, and if any good Semiticist can see the obvious derivation from a mimated form of Semitic *sabˁ- 'seven', then there's no denying the cultural contact (unless one has another agenda other than straight facts, that is).

    In order to place Semitic in the deep south, PIE must then be placed in Turkey to address the IE-Semitic contact. Yet the Out-Of-Anatolia theory is impossible and the consensus is secure in placing PIE in the NW Pontic region for several thorough reasons.

    So the conclusion is clear: Proto-Semites were agriculturalists living to the north as far as Syria and Turkey. Semites need not be any further north than that since I reckon that sea trade can cleanly account for the rest.

    I see no valid reasons argued against my idea despite much opinionated quibbling about the ur-locale of Semitic in literature and online. I want to deal in facts, not empty opinions of one or another.

    ReplyDelete
  7. Quick question... Though this is a bit off topic, what are your thoughts on the Urheimat of the Afro-Asiatic language family?

    ReplyDelete
  8. I think the distribution of AA suggests an origin from East Africa. However within the larger context of the Nostratic hypothesis, AA must then have first entered Africa from the Near East before spreading out.

    Many people online banter back and forth with genetics and archaeology to prove their claims of language movement but I've said many times that these considerations are only weakly connected at best. More detailed etymological work has to take place before we can be sure at all of any linguistic changes that took place such a long time ago (potentially about 10,000 years ago or more).

    ReplyDelete
    Replies
    1. "I think the distribution of AA suggests an origin from East Africa. However within the larger context of the Nostratic hypothesis, AA must then have first entered Africa from the Near East before spreading out."

      An East African (or southeast Saharan) origin seems probable IMO, but I will always be bothered by these apparent "loanwords from the Sumerian and Caucasian languages" within branches of AA. This would give some credence to the Nostratic hypothesis, though debate flourishes as to whether AA should be included into Nostratic. Not to mention the problem of whether these terms can be successfully reconstructed to Proto-AA, or whether they were introduced with the later spread of West Asian domesticated animals into Northern Africa. (I need to read up more on this.)

      "Many people online banter back and forth with genetics and archaeology to prove their claims of language movement but I've said many times that these considerations are only weakly connected at best."

      Yes, I've read through many of the arguments on forums like www.forumbiodiversity.com and s1.zetaboards.com, but some often seem to be argued with an ethnocentric bias from both sides of the issue. Haplogroup/haplotype (mitochondrial, Y-chromosomal) and autosomal genetics, along with archaeology, provide substantial bases for the claims made, but as you said, the considerations are only weakly connected.

      However, I will always be interested in the fact that Y-chromosomal haplogroups E1b1b, J1, and R1b1a (marker: R-V88) are found on both sides of the Red Sea, albeit at varying frequencies. Whether or not these haplogroups can be associated with AA dispersal, they do make you raise an eyebrow.

      "More detailed etymological work has to take place before we can be sure at all of any linguistic changes that took place such a long time ago (potentially about 10,000 years ago or more)."

      Agreed.

      Delete
  9. Where I can find a checklist of IE-Semitic loanwords? I know about *gHaido- "goat", *tauro- "bull", h2ster- "star", septm "seven". PIE loans from PSem, or from a third source, common to both?

    ReplyDelete