One of the hypotheses supported by Theo Vennemann and other linguists is that Proto-Germanic was influenced by some Semitic language. The evidence they present for their case includes:

  • Loss of some grammatical cases from Proto-Indo-European. This would be an indication of language contact, with adults trying to learn a different language and simplifying it in the process. This phenomenon has not occured in other languages spoken at the time, such as Latin and Greek.
  • One third of Germanic roots do not trace back to PIE, and some of these words seem to have common roots with Semitic languages. For example, Proto-Germanic *furkhtaz, Proto-Semitic *prkh, 'fright'; Proto-Germanic *magaþ, Early Semitic makhat, 'maiden'.
  • Grimm’s law, that has introduced the fricative consonants *[f], *[h] and *[θ]. PIE was poor in fricatives, compared to Semitic languages.
  • Some deity names also seem to have common origins, such as Old High German Phol and Semitic Baal. These names can also be derived independently through regular sound changes, such as the Grimm’s law.
  • Verbs are inflected for tense only in the present and past (like Semitic languages). Other Indo-European languages have a much richer system of verb inflections for marking tense.
  • The use of ablaut for marking the past of strong verbs.

Are these observations strong enough to posit a Semitic substrate in Proto-Germanic?

  • 5
    Dear downvoter, could you please explain me what is wrong with the question? – Otavio Macedo Oct 30 '11 at 13:30
  • 2
    I didn't vote your question down, I think it's interesting. I believe Vennemann's views are still highly controversial, but you probably know that. – Cerberus Oct 30 '11 at 18:22
  • 2
    Venneman doesn't claim it to have been Phoenician (the time depth he's talking about is around 5000 BC, so it predates Phoenician). He calls the group the 'Atlantic' peoples, and claims they spoke a Semitic language. – Gaston Ümlaut Oct 31 '11 at 3:24
  • @GastonÜmlaut, fair enough. Question edited. – Otavio Macedo Oct 31 '11 at 9:01
  • 2
    I can't say I like this question because it completely covers the answer, leaving the only option for other answers is to post a rebuttal which the asker can then accept. I'm tempted to copy the answer embedded in your question to an answer so that people have an opportunity to vote. Also, the rebuttal answer is at it's heart a "well, we'll never know" sort of answer, which is the more difficult to rebut. Anyone can advance skepticism as an answer to scientific inquiry. – MatthewMartin Nov 3 '11 at 16:01
up vote 17 down vote accepted

This seems like a weak case. I'm sure there's a much more thorough rebuttal in the literature, but I'll give at least a few contrary remarks.

The second point, that some words in the two languages seem to have some sounds in common, is the most frequent unscientific argument presented in linguistics circles for the existence of a language family. The problem is that between any two sufficiently large sets of data (language vocabularies have thousands of items), there will nearly always be a couple dozen pairs of words that can be extracted that are "similar enough" to be cognates. But that isn't how sound change works - to demonstrate shared ancestry, you must reconstruct words of the common ancestor language, and give a series of regular sound changes that output words in the daughter languages - Grimm's law is a model example of this.

The late Basque linguist Larry Trask gave a quite readable defense against the ceaseless proposals attempting to connect Basque, a language isolate, with nearly every other language family, in "Origin and Relatives of the Basque Language" (1995), which I highly recommend for its applicability to this case.

I also fail to see how a Proto-Semitic influence would have induced PIE to replace *[k] and *[t] with *[x] and *[θ] when both languages had *[k] and *[t]. Furthermore, PS doesn't even have an *[f] sound, so PIE could not have borrowed it. (You gave *[h], instead of *[x], as the reflex of PIE *[k] - this was a later development in English).

Proto-Indo-European had only two tenses, present and past, on verbs in the imperfective aspect. Tense on other verbs was unmarked. The rich tense systems of its descendants are modern innovations.

"Ablaut" refers to a morphological alternation already present in PIE - you're thinking of Germanic umlaut, a process which is uncontroversially understood as the product of fronting a stem vowel before a suffix containing [i] (the suffix is later dropped). A substrate influence is not needed to explain this.

That being said, you're right that the loss of complexity in the case marking system and the large number of words unique to Proto-Germanic may indicate a pre-IE linguistic substrate. Unfortunately, barring a revolutionary discovery of a new trove of data, we will most likely never know.

  • 3
    Here's a fairly detailed rebuttal of Venneman's claim for Vasconic and Semitic influence in early Europe, as made in his 2003 book 'Europa Vasconica - Europa Semitica'. – Gaston Ümlaut Oct 31 '11 at 3:28
  • 1
    No, he does mean Ablaut. Germanic languages are notable for having partly regular patterns of sound changes in their vowel paradigms, which go back to PIE ablaut. These are sometimes present in other branches, but they're not so obvious, so it's thought of as a distinctive feature of Germanic – Colin Fine Dec 7 '13 at 0:25
  • Sorry, that can't be right. The Germanic alternation is phonetically motivated: assimilation by fronting of a back vowel when followed by another front vowel (in this case, *[i]). The PIE alternation is unpredictable (well, depending on whether you've done internal reconstruction to recover the laryngeal series, but that's beside the point). Unpredictable alternations do not suddenly pick up phonetic conditions. – Alek Storm Dec 9 '13 at 2:27
  • The first part of your answer seems off-topic. There was no mention of Basque in the OP's question, and Vennemann doesn't posit a genetic relationship between Vasconic, Semitic and Germanic, but one of sub/superstrate influence respectively. As for ablaut, there are several internally regular series (with exceptions of course). But I don't see where a phonetic condition comes into play? Neither the OP nor Colin Fine mentions that. – Marc Schütz Nov 17 '16 at 14:07

I think you couldbe right... I also have thought about this. I myself found that there are a lot of words and names that have a similar sound and also a similar meaning. could be coincidence, but i compared the (seemingly semitic) german words with the ancient hebrew words and also with some latin languages and in many cases the latin languages didn't copy all the semitic words that the germans did copy. one example: the verb calling/shouting (there are different version of this word, both in hebrew as in the germanic languages.

Modern Germanic languages: 1. Dutch: roepen. Danish: ringe. German: rufen. Norwegian: ring. 2. Dutch: schreeuwen. Danish: rabe. German: schreien. Norwegian: rope. 3. Dutch: brullen. German: röhren. Danish: brol. Norwegian: roar. Eng: roar.

Old Hebrew: 1. רוע Ruwa/Roa = Shouting/calling 2. רנה Rinna = giving a shout/calling out See the bibles verses: Jozua 6:16 & Psalm 17:1 for an example of those words.

Latin: clamor & vocatio.

Scientifically this may not be usable, especially because in this example I used the modern words. Nevertheless, also looking at the Y-DNA profiles of the german men there is a real mixture visable between two groups. One which is very comon in the western latin european countries (R1b) and is present in about 50-90% of the men and another group (I1) which is less comon (20-30%) mainly found in northern europe and scandinavia. The I1 group is most closely related to what today are the semitic people (group J: arabs, jews etc.) and are believed to have migrated already a while ago from the area of israel/iraq/syria etc. to europe. the R1b group has it's origins more in the area of eastern iran/afghanistan etc. and could have always spoken a indo-european language. The I1 group however could have originately spoken a semitic language before they mixed with the R1b's. Because the R1b group is much bigger nowadays (I1 is in no place the dominant group) it seems logical that the I1's slowly adopted the indo-europen language as they mixed in. However we still find some traces in the german languages that remind us of the semitic heritage of part of the germans.

That's my theory ;)

  • 2
    I can't make any sense of what you are saying in your Modern Germanic languages (I can't even work out how many distinct roots you are discussing), but none of them match the consonants of either of the Hebrew roots you mention. – Colin Fine Dec 7 '13 at 0:29

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.