1. The WordReference Forums have moved to new forum software. (Details)

French: Gaulish substratum theory

Discussion in 'Etymology, History of languages and Linguistics (EHL)' started by Cilquiestsuens, Sep 13, 2013.

  1. Cilquiestsuens Senior Member

    Moderator note: Split from here.

    Well, the problem with words like carrus, is that Le Petit Robert gives it as deriving from latin carrus, with no mention whatsoever of Gaulish. Its Gaulish etymology might be 'widely accepted' among diachronic linguists, but even educated people from other fields wouldn't know about that.

    In any case, I have found the link to the said article: here, it is called: ''Pour une réévaluation du substrat celtique et pré-indoeuropéen du lexique français'' by Gilles Quentel, Université de Gdansk.

    My query was about 'chien' and I have read your interesting link, that could explain the change historically, since the shift seems to be accepted as a pattern observed in other words.

    As a linguist, do you think it is worth give a thought to that substratum theory?
    Last edited: Sep 13, 2013
  2. berndf Moderator

    German (Germany)
    The fact that the Latin carrus is itself a Gaulish loan does not alter the fact that the French char is inherited from Latin. Whatever merits the the Gaulish substratum theory for French might have, char is not a valid example.
  3. CapnPrep Senior Member

    Educated people who are interested in etymology understand that dictionaries like the Robert or Petit Robert usually only give the immediate origin (Latin in this case) and have to leave out most of the details. They also know that Latin words themselves have an etymology (and that a dictionary of French like the Robert is not the right place to look for that).
  4. Cilquiestsuens Senior Member

    As a linguist, do you think it is worth give a thought to that substratum theory? What are the pros and the cons?
  5. CapnPrep Senior Member

    I think that most linguists agree that French has elements of a Celtic substrate. We know very little about Gaulish, so many claims about its influence on French remain speculative. There's nothing wrong with speculating, but I find many of Quentel's examples to be poorly argued and misinformed about existing work, and the suggestion that there has been some kind of conspiracy to minimize Gaulish/Celtic etymologies in French in favor of Latin borders on crackpot territory. :(
  6. fdb Senior Member

    Cambridge, UK
    French (France)
    I agree totally.
  7. Cilquiestsuens Senior Member

    Moderator note: From the original thread (here).

    I thought you meant to say I had said that.

    Whatever Mr. Quentel's article's worth, I was interested by two things in it:

    1. He says that the Gaulish substratum has been underestimated. I would agree and I don't think he mentions any conspiracy. The reason for this state of affairs is that Celtic languages are the underdogs of linguistic studies (No offense meant to linguists, this is just the logical outcome of the Celts' history and their present condition. They had their days of glory long long ago, I guess). Just out of curiosity: How many Celtic languages do you, professional linguists, know fluently?

    2. I thought the idea of the double negation (ne...pas) of French being due to substratum influence is interesting. What is unfortunate with Mr. Quentel's research on that is very scanty and far from being able to lead to any conclusion.

    Overall he touches on a number of interesting (but a bit random) facts without presenting any research or data to back up his claims. But that, in itself, doesn't disprove his theories. Overall his article looks more like the final draft of a quick brainstorming session.

    Last edited: Sep 13, 2013
  8. berndf Moderator

    German (Germany)
    This is about Gaulish and not about Insular Celtic languages. It would be impossible to "know them fluently". We only know a few words of Gaulish from mainly Roman sources and from toponyms* that have survived. The Celtic languages of the continent became extinct before anybody developed a writing system system for them. The Druidic culture was very much based on privileged knowledge they jealously guarded to themselves and this once dominant culture of central and Western Europe never left anything written. We even don't know how similar Insular and Continental Celtic really were.
    * E.g. we know that the root hal must have meant salt from all the place names Reichenhall, Hallein, Hallstatt in the area around modern Salzburg, a region famous for salt mining and the cradle of the Celtic culture.
  9. CapnPrep Senior Member

    On the contrary, Celtic languages have received a tremendous amount of attention from philologists, precisely because of their history (in particular because of the textual records and other linguistic evidence they have left). The Celtic peoples' present condition doesn't really have much effect — either positively or negatively — on etymologists…
    I thought this statement was interesting: "Une chose est certaine : aucune autre langue indoeuropéenne, pas même le latin, n’utilise systématiquement la négation en deux parties." The same structure can be found in varieties of Occitan and Catalan, and I believe in some Italian dialects, and the Latin origin of the reinforcing elements (pas, point, mie, goutte, etc.) presents no phonetic or semantic difficulties. Bipartite negation is also found in Germanic, for example in the history of English and German.
  10. Cilquiestsuens Senior Member

    You are just confirming what I said about Mr. Quentel and his quoted paper.

    You wrote: ''I believe in some Italian dialects''. Which dialects? When? If you are talking about Northern dialects, then we never left the Gaulish linguistic area (covering Catalan, Occitan, Provençal and Oïl languages and including some Germanic territory too). To dismiss this theory shouldn't we rather look at other areas outside of Celtic influence?

    Or maybe you think that this double negation is a much later addition to the language? When did that occur?
    Last edited: Sep 14, 2013
  11. Cilquiestsuens Senior Member

    So you claim that knowledge of 'Insular' Celtic languages is completely useless !!! There is a simple fact you overlook, the knowledge of Continental Celtic is so scanty that the only other (secondary but not 'useless') source are Insular Celtic languages.

    In the example you give above (hal = salt), in Modern Breton salt is still Holen. It illustrates how the Celtic vocabulary hasn't changed much over centuries and the strata of borrowing are well charted (1. Latin 2. Middle French / English)

    Breton, and this might be actually the most interesting point raised by Mr. Quentel, and its numerous dialects, due to its historical position, is the surviving Celtic language which has the closest relationship to Gaulish. It is thought that some Gaulish element (vocabulary, syntax, phonology) have survived in some Breton dialects; there were heated arguments between specialists on this topic. It was initially thought that the dialect of Gwened (Vannetais) was the one having maintained a Gaulish element, this was François Falc’hun’s theory, but Kenneth H. Jackson tried his best to discard it. Another expert in this field was Léon Fleuriot.

    So according to you all those eminent Celtic linguists were crackpots?
    Last edited: Sep 14, 2013
  12. CapnPrep Senior Member

    Not before the 12th century, and definitely long after Gaulish had died out. And it appears at around the same time in Occitan, so there is no evidence of gradual diffusion from areas in contact with Breton.
  13. JeanDeSponde

    JeanDeSponde Senior Member

    France, Lyon area
    France, Français
    Indeed he didn't say "there is a conspiracy."
    Yet many sentences tend to cast a shadow over the honesty, impartiality or competence of philologists:
    His general idea is clearly "I wonder why philologists refuse to consider non-latin substratum".
    And note that all his examples are in the line of "why did they deny a possible Breton origin?".
  14. berndf Moderator

    German (Germany)
    Falc’hun obviously had a political (French nationalist and anti-separatist) agenda that casts serious doubts on his scientific objectivity. Fleuriot, to my knowledge, never admitted more that a possibility of some Aremorican influence on Breton.That is exactly the problem: We simply don't know how similar the groups really were. This all remains dreadfully speculative.
  15. Cilquiestsuens Senior Member

    If indeed this double negation appears when you say it did, the argument of a Gaulish substratum having some kind of influence in that process should be completely ruled out... (except if that was a much older phenomenon that started being recorded in written very late?)

    Do you have any good link to suggest about the historic appearance of this phenomenon in the languages affected?
    Last edited: Sep 14, 2013
  16. Cilquiestsuens Senior Member

    A Breton origin???Who said that?? How is that possible ? Are you misquoting?

    Breton here is mentioned for possible parallels deriving from a common heritage. No one ever spoke of Breton origins... The whole thing is about Gaulish origins. Are you following?
  17. Cilquiestsuens Senior Member

    You're making a case against Falc'hun who was disliked by all the extremists and crackpot elements of the Breton movement.

    He had the respect of the moderate ones because he was an eminent linguist and Fleuriot had accepted a number of his views relating to the survival of Old Celtic (pre-Insular) forms in Breton dialects. What he never accepted was that Vannetais (Guenedeg or Gwenedeg in peurunvan spelling) was the descendant of Gaulish.
    Last edited: Sep 14, 2013
  18. JeanDeSponde

    JeanDeSponde Senior Member

    France, Lyon area
    France, Français
    Here are a few "I wonder about breton" from M. Quentel:
    Can you name another dialect / language as abundantly quoted in the article as breton...?

    I'm not a philologist (by far). I'm a (hard) scientist. When opponents to Darwin say, "There is no proof that an eye could naturally evolve from some light-sensitive cell", I can just say "where are you speaking from...?"
    Last edited: Sep 14, 2013
  19. Cilquiestsuens Senior Member

    Is there anything wrong about Breton being the most quoted language in the article ???

    I quote your post above (#13):

    The word in contention here was ''Breton origin'' not ''Breton''. In your previous post you were not able to prove that the author of the article was saying what you wrongly attributed to him ("why did they deny a possible Breton origin?").

    As you should have understood, the point of quoting Breton is that it is the surviving language which the closest relationship to Gaulish. It is not about Breton origins, but parallels. What might be confusing, is that some of the parallels quoted in the article seem to be a tad far-fetched and extremely debatable or even pointless (such as 'e-pad' as a possible explanation of 'pendant')
  20. JeanDeSponde

    JeanDeSponde Senior Member

    France, Lyon area
    France, Français
    Please be consistent :
  21. Cilquiestsuens Senior Member

    I don't understand your misunderstanding, but I understand it is persisting.

    Please, get back to the main topic and set those trifles aside.
  22. CapnPrep Senior Member

    It will be discussed in any grammar of Old French or Old Occitan/Provençal, etc. The chronology has also been mentioned in a couple of threads here on WR:
    Why were negative reinforcers used in Old French
    FR: Qui ne dit mot consent
    As berndf already pointed out, it is impossible to say how similar Breton and Gaulish were, so Quentel's suggestion that examples from Breton (especially modern Breton) can be taken as a kind of stand-in for Gaulish is really hard to take seriously.
  23. JeanDeSponde

    JeanDeSponde Senior Member

    France, Lyon area
    France, Français
    Practical "Gaulish substratum" examples mentioned by Quentel are taken from (modern) breton, this language being supposed as a kind of "intermediary though authentic" medium.
    As you fail to admit it, I'm unsuscribing from this thread.
  24. Kentel Junior Member

    I read a few weeks ago this very interesting discussion about this paper "Pour une réévaluation du substrat celtique etc.", which, as Cilquiestsuens pointed out, is a kind of "final draft after a quick brainstorming session" which occured during a conference I attended some years ago.

    You are right to emphasize its methodological flaws : there is no method there in fact, and when I wrote it my idea was more to elaborate a blueprint, the general axis of my new research and to spot the problems revolving around it. Thus,there is no "discovery" there, no conclusion, no new method and I didn't have any political agenda either. I have written more interesting (at least I hope) papers since. For copyright reasons, I couldn't put them on my page, but if some of you are interested I'll gladly send them some copies.

    I agree with several remarks made by contributors here (Capnprep especially), and Cilquiestsuens has very well answered to many objections already, and he's also right to say that all this has been done "a bit a random". So, to put it all back in order, here are the general ideas underlying this paper - and all my research so far.

    1- Substrata : my claim is, in a nutshell, that the influence of substrata is undervalued. Not only in French, in all the IE languages. I have been working especially on Romance, Germanic and Celtic, but the problem is true everywhere else. There has been, in historical linguistics,a myth, which says that a new language imposed by conquerors eradicates completely the language(s) spoken by the indigeneous population(s).

    This goes not only for French (= the invasion of Gaul by the Romans and the eradication of Celtic), but also for English, Scandinavian, Spanish, Sanskrit and it is especially true for the alleged Indo-European conquest of Europe. No substratic influence remains. My first statement was that it couldn't be true, because it is not realistic.

    Take French (I guess most of you here can speak french since you were able to read my paper): French speakers have specific articulatory habitudes. When you listen to a Frenchman speaking English, you can hear that the phonetic distorsions are huge, that he/she uses syntactic calques all the time. In other words, that his/her use of the English language is heavily distorted by the influence of his mother tongue. When the Romans conquered Gaul, the same process certainly occured, as well as it occurred in India when the Aryan population settled upon the Dravidian one, as well as it occurred in England when the Germans settled in Celtic speaking areas, and so on.

    2- Innovation vs. Inheritance
    : historical linguistics' mainstream theory claims that all these phonetic/morphologic/syntactic changes are the products of later evolution. For example, we are supposed to use nasal vowels in French because the sequence vowel + nasal consonant yields a nasal vowel. This process is called innovation. You get a new sound which didn't exist previously - or a new grammatical feature (the article, the compound tenses, the -bo future in Latin or the internal past structure of the Germanic verbs f.ex.). I don't believe in innovation, I believe in inheritance, which is the exact opposite. We have nasal vowels in French because there has always been nasal vowels in the area (eg Northern France). Innovation is the product of structuralism : the system evolves independantly from any external influence, by the random interactions between its internal structures. This has led to many false conclusions in historical linguistics and in several other fields (see Dumézil's theory f.ex).

    Hence we have a debate inheritance vs innovation. But innovation is in most cases unrealistic; if you take Mycenean Greek and Modern Greek, you have the longest observable time span of linguistic evolution in Europe (3500 years), and you can observe that the words have very little evolved phonetically. So, why believe that from Latin to French (ie. 1000 years), the evolution was so huge and so sudden ? Why believe that from PIE (3000 BC) to Mycenean Greek so much changes occured in 1500 years only, and that afterwards everything went so slowly during 3500 years ?

    For me, the explanation is: inheritance of the ancient substrata. Greek did not evolve because from 1700 BC to modern time, Greece has been (more or less) populated only by greek-speaking populations. Therefore the slow evolution: there is no substratal influence after 1800 BC. On the contrary, Latin was spoken until the 5th century AD on Celtic (and Iberian, Thracian etc) speaking areas, which explains the considerable and sudden amount of alterations in both morphology and phonetics. And the lexicon has undergone the same substratic influences.

    I don't deny that innovation played its part in the process, but inheritance was probably the key element of the story. And my research is about that. Now if you have questions about more specific points, or remarks or critics, I'll try to answer them :)
  25. Kevin Beach

    Kevin Beach Senior Member

    What fascinates me about etymology generally is that one can find ancient secrets "hidden" right in the open. As in comparing North-east English speech rhythms with those of Scandinavia, or listening to (mainly rural) North American accents and spotting the British/Irish regional accents that sired them.

    When considering the substrata of Romance languages, it's also valid to look at what Italic languages apart from Latin may have been prevalent in and around Italy before Roman supremacy took hold. The gradual conquest of other regions took Latin to places where similar languages were spoken. The imposition of Latin by the Roman conquerors would have caused a melding of similar tongues (as with Old English and Danish in the 10th and 11th centuries in England), creating the varieties that were left after the Roman empire fell. French pronunciation may well be the product of that process as well as of the influence of non-Italic languages.
  26. Cilquiestsuens Senior Member

    Many thanks, Aotrou Kentel, ha mersi bras deoc'h for taking part in this discussion.

    The obvious question that comes to my mind is : were you able to make any further research in this field and have you found anything that might substantiate your assertion about the underestimation of the Gaulish substratum? In other words, are they any words about which it can comfortably be asserted now that their traditionally held Latin origin should be discarded for a Gaulish one?

    I am also curious to know about the nasal vowels of French. Can any kind of research prove anything in that matter?
  27. Kentel Junior Member

    Mann ebet, plijout a ra din komz war-benn yezhoniezh gant tud a zo dedennet gant an traoù-se:)

    As a matter of fact yes; I compiled a list of words which etymologies are unlikely. There are many. However, pointing at absurd etymologies is not enough, you have to propose new explanations or new methods to deal with them, which implies invalidating the previous ones.

    I have written two papers specifically devoted to the substratic origins of some French words; in one of them I have first tried to demonstrate that the echoic theory (= you make a word out of a sound, like "tomber" imitating the sound of a fall) was a nonsense. Once you have invalidated and etymological process (echoism here), then many words are left without etymologies, and there you can propose something new - which I did for the following words :

    - petit, from Gaulish "pettia" and not from an expressive root "pitt-", pépin has the same origins.
    - coq (rooster), from Gaulish "caliaco" and not from the sound of its cry. The etymology of coquelicot is connected to this one.
    - marmotte (marmot) and marmonner (mumble), probably from Gaulish marunata "dirge", and not from the sound of... of what by the way ? we don't know, but the word is allegedly "echoic".
    - murmurer (to whisper), certainly from the same root and not from the Latin "murmur" which has exactly the contrary meaning (to rumble).
    - pic (peak), from Gaulish "picos" (peak), and not from an expressive root.
    - pinson, from Gaulish "pincio" (same meaning, ie. "chaffinch") and not from Vulgar Latin *pincionem
    - traquer (to hunt down), from Gaulish "tragula" (track), itself from "trago" (to walk) and itself from tragos (foot). Trace is certainly connected, as well is obviously the English "track".
    - vanneau (peewit), from Proto-Celtic *vanello-, and not from Latin vannere (the same statement have been made by Pierre-Yves Lambert in "La Langue Gauloise - Errance 2003", as well as for "petit").

    As for pre-Celtic substratic origins, I have listed the following words (but I have many others at hand): canard, tomber and toucher. You can add safely trouver and tuer, at least. I've published these results in the Revue Romane (at Benjamins) in a paper titled "La Néologie Expressive et onomatopéique dans l'étymologie du français".

    To put it fast, I have written another one "Le Substrat Celtique dans l'étymologie du français" which publication is scheduled for spring 2015, with many other words.

    I have also written a paper about the common pre-IE lexicon of Celtic and Germanic, and another one focusing on the specific pre-IE lexicon of Scandinavian (not shared with Western nor Eastern Germanic). And I have proposed a new method in order to check these lexical items.

    Right now I'm working a book on the proto-IE origins of the verb in Celtic, and this book will have highly substratic-friendly contents, if I can put it this way - especially as far as the compound tenses are concerned. I am also working on a paper about the substratic origins of the postponed article in Scandinavian and about an etymological device named "endocentric composition" which I will try to discard. I stop there, my post is already too long I fear...
    Last edited: Oct 1, 2013
  28. Kentel Junior Member

    This is a very complicated matter. Little research has been made about that recently as far as I know. Historical linguistics is an almost dead field of research now unfortunately. Anyway, the common view on the nasals is structuralist : oral vowel + nasal consonant = nasal vowel, under certain conditions.

    What I think is pretty different : oral vowel + nasal consonant does not yield a nasal vowel in Spanish nor in Italian: you have FR "penser" ([pãse]) but ES "pensar" ([pensar]). Thus I ask the question which is never asked : why a nasal vowel in French and not in Spanish nor in Italian ? Structuralists have no answer to that. Or yes, they have: it's the result of internal evolution, by dissimilations and assimilations (you don't know exactly which ones and you dont' want to ask).

    For me, whichever are the internal evolutions, the dissimilations and the assimilations, a language does not create a completely new set of sounds out of the blue. If in Northern France people began to use nasals, while they did not in Southern France nor in Italy nor in Spain, there must be a good reason. And if you think about it, there are not many possible answers...

    The Celtic influence has been proposed, because you have many nasals in Breton and in Gaelic, and because Portuguese, (which has possibly a Celtic substratum), also has nasal vowels. But I am not so sure : we cannot demonstrate so far the presence of nasals in Gaulish, and there's no nasals in Welsh nor in Cornish - not even in Old Welsh. Hence it's difficult to claim that nasal vowels are typically Celtic. I think it's older, it goes back to pre-IE languages.

    The same question could be ask for the French "r" and "u", by the way.
    Last edited: Oct 1, 2013
  29. Kentel Junior Member

    That's a very good question indeed, because it underlies the problem of proto-Romance. We know that the Romance languages do not come from classical Latin, but from a heavily altered form of Latin named Vulgar Latin. We have no traces of this language though, and we don't know if it is a sister or a daughter of classical Latin.

    Among the most striking differences between Classical and Vulgar Latin, you have a great part of the lexicon which makes an extensive use of diminutive affixes (solicullum, avicellum, instead of solem and avis, f.ex), and many words of non-Latin origins. You have also the compound tenses and the articles, which do not exist in Classical Latin, you have no -bo future nor -bam imperfect, probably no morphological passive either, etc.

    The influence of the neighbouring languages could have played its part, but it is hard to tell which one: the sabellic languages have been absorbed by Latin at a very early stage, and they are very archaic (= close to PIE). You have traces of Oscan in Latin, like f.ex. the word "bos" (the outcome of PIE *gwos should have been "vos" in Latin).

    Etruscan certainly has influenced Latin, but it's difficult to measure this influence since very little is known of it. The lexicon especially, is almost unknown : most of the epigraphic documentation is extremely repetitive, and although we have a lot of inscriptions they say almost the same thing with the same words all the time.

    Etruscan and the Sabellic languages played their part at a very early stage during the formation of Latin. Did they influenced Vulgar Latin too is impossible to tell, but until now it seems quite unlikely. The most realistic is that the Roman soldiers who conquered Gaul spoke a relatively homogeneous lingua franca called Vulgar Latin.

    But the question of the origins of Vulgar Latin still remains, and therefore remains the question of the substratic influence of Sabellic, Etruscan, Illyrian, pre-IE on this language.
    Last edited: Oct 1, 2013
  30. Cilquiestsuens Senior Member

    Plijadur vras am eus bet o lenn ho respontoù, ha trugarez dezho ha deoc'h em eus bet tro da zeskiñ ur bern traoù dedennus-kenañ!

    Getting back to the main topic, I had a few questions to start with about two facts I had taken so far for granted.

    1. Is the partly vigesimal counting system of French attributed to Celtic influence?

    2. Another example: the word 'temps' (undoubtedly Latin) means both 'time' and 'weather'. Don't we have here an example of Celtic semantics in Latin attire ??? (Cf. Breton amzer, Irish aimsir)

    To be honest with you I have been extremely interested by all your examples, including the ones about Pre-Indo-European (very fascinating).

    Would it be possible for you to elaborate more about the word petit and how you managed to discard its alleged onomatopoeic etymology and how you then connected it to the Gaulish word pettia (a 'piece' or 'chunk', which Breton speakers will recognize in the word pezh) ? And by the way, isn't the French word pièce likely to come from pettia too???
    Last edited: Oct 7, 2013
  31. fdb Senior Member

    Cambridge, UK
    French (France)
    Italian and Portuguese tempo, Spanish tiempo also mean “weather”. Russian время is “time” and “weather”, and similarly in other Slavic languages. Modern Greek καιρός again has both meanings.
  32. Cilquiestsuens Senior Member

    Thank you very much for your input! So it looks more like pan-Indo-European, I guess. I still have two questions.

    1. It seems to me Latin didn't have this feature? Nor Greek? Sanskrit? Avestan?
    2. Do we know of any ancient Indo-European language having this feature or if it is not the case, that could be thought of as a something that might be pre-Indo-European?
  33. CapnPrep Senior Member

  34. Cilquiestsuens Senior Member

    Thank you very much for these very interesting and eye-opening links.

    What about the vigesimal system?

    What do we know about it concerning old languages? We know it is widespread in Modern Celtic languages but do we know anything about its usage in antiquity? Any hints at a vigesimal system in Gaulish?

    It seems it exists in Basque too.
  35. Ben Jamin Senior Member

    But not in Polish.
  36. CapnPrep Senior Member

    Here's a quote from Vennemann:
    Vennemann of course thinks the origin is Basque.
  37. berndf Moderator

    German (Germany)
    Quatre-vingt(s) didn't occur in French before the 12th century. Wherever it came from, it doesn't appear to belong to a Gaulish substratum inherited directly from Proto-French but either to be a loan translation or an independent innovation in Old French. It might be taken from a Celtic language but it might well also be of Normanic origin as it probably is in English (e.g. threescore and ten for 70 in the KJV of the Bible).
  38. Kentel Junior Member

    This is a very good question (I don't say that rhetorically). You have indeed a 20 base counting system in Celtic and Basque, probably in Gaulish too (uoconti = 20, see Delamarre's dictionary), from which French has inherited its own counting system. You find it in many languages in fact (Georgian, Persian, Maya and many North Amerindian languages including inuit). Even the latin "viginta" (possibly a borrowing from Celtic), so it's not uncommon.

    Here is a little material for thinking :

    The Danish counting system is the only one with a 20 base counting system in Germanic :

    ti = 10
    tyve = 20 (note that English has a 10 base system)
    tretti = tre (3) x ti (10)
    halvtreds = 50, litteraly 3 x 20 - (20/2), or halvtredsindstyve ("the half part of twenty substracted to three times twenty)
    treds = 60, litteraly 3 x 20, or tredsindstyve (three times twenty)
    halvfirs = 70, litteraly 4 x 20 - (20/2), or halvfirsindstyve (see 50)
    firs = 80, litteraly 4 x 20
    halvfems = 90, litteraly 5 x 20 - (20/2) or halvfemsindstyve
    hundrede = 100

    Compare with Norwegian :

    ti = 10
    tyve = 20
    tretti = 30
    førti = 40
    femti = 50
    seksti = 60
    sytti = 70
    otti = 80
    nitti = 90

    Another mind-challenging one :

    Breton :

    6 = c'hwec'h
    18 = tri-c'hwec'h (so-called "senary counting system", like the one on a clock)

    Another one :

    Gaelic :

    1= ceann ("head"), or "a haon" for the classic version

    1 people = duine (= "person")
    2 people = beirt

    Polish :

    1 = jeden (classic) but also "raz", which is more common in many situations.

    Note also the discrepancies between cardinal and ordinal counting systems for the two first numbers :

    1= one/first/premier + Breton kentañ + Scandinavian først
    2 = two/second + Scandinavian anden

    I don't claim that everything there is substratic; there is an agreement to consider the Danish counting system as very archaic, but how archaic is the question. Numbering was probably invented with trade, or a least with farming, not before (many aborigenal populations have no counting systems at all, or only 1-2).

    There is another matter for thought is you want to investigate substratic influences : the names of the four cardinal points and the names for right and left (especially left...).
  39. Kentel Junior Member

    Innovation would mean in this case that people suddenly decided not only to change the system, but also to use a more complex one. Why ?

    If a loan, a loan from which language ? Old Icelandic has a 10 base counting system.

    the XIIth century sounds a very early date to me : 20 itself is only attested at the end of the XIth. There is much to say about the counting system in Old French, by the way.
  40. berndf Moderator

    German (Germany)
    What about Old, Middle and Early Modern English? The origin of the score-based counting system is clearly Nordic.
    Mentioned this mainly for completeness. I assume it is a loan translation.
    See above in this quote.
    Attestation mentioned here. The Dicionnaire de l'académie francaise says the same.
    Last edited: Oct 8, 2013
  41. fdb Senior Member

    Cambridge, UK
    French (France)
    Persian no, Ossetic yes (presumably from the Caucasian substratum)

    Other than English ("three score years and ten"). Since English "score" is clearly a Norse loanword it is not surprising to see the same phenomenon in Danish.

    PS.: Overlap with Bernd.
  42. Kentel Junior Member

    Of course.
  43. Kentel Junior Member

    You got the point :) I don't know much of Persian, I confess.

    It is in my opinion, because you don't have it in old Norse; the score = 20 is very rare, I have never seen it in sagas (I confess I havn't read them all - far from it) nor in Beowulf as far as I can remember. But yes, you're right, the word is borrowed from Scandinavian; the 20 base system does not appear to be Germanic whatsoever, because most Germanic languages do not use it.
    Last edited: Oct 8, 2013
  44. CapnPrep Senior Member

    What is your evidence for saying "probably in Gaulish too"? The form of the word "20" tells us nothing (in fact, if it is derived from roots meaning "double ten", it is a decimal form…) There is no indication that it was used in the formation of larger numbers, and as mentioned above, there is an attested form for "30" that is clearly decimal.
  45. berndf Moderator

    German (Germany)
    There is an uninterrupted history of 20-based counting in English that goes all the way to the early 20th century where the word hundred is "defined" in the OED as fivescore.
  46. Kentel Junior Member

    If not from Gaulish, from where ? Old Norse has an attested 10 base system (with "score", I got it). Celtic has an attested 20 base system. And Basque (let's be Vennemanians).

    The form of the word is pointless in this debate, the point is the system : how you count. By saying quatre-vingts or firs, you imply 4x20. Only that matters. "Score" means "mark" in PGmc, and it's PIE too, and it is adapted to express 20.
  47. berndf Moderator

    German (Germany)
    I am not so sure:
    Score is not only a word but the base of a counting system. Threescore is another word for sixty, fourscore is another word for eighty and 70 can be pronounced threescore and ten (Bible) and 90 fourscore and ten (A.Lincoln). The 10 and the 20-based words were used side by side, like octante, huitante and quatre-vingts in French.
    Last edited: Oct 8, 2013
  48. CapnPrep Senior Member

    That was my point: How did they count in Gaulish, by tens or by twenties? They had a decimal form for "30"; for higher numbers, we have no idea, do we?
  49. Kentel Junior Member

    And what does it tell you about the origins of this 20 base counting system in English ?

    Back to the subject: we have stated that

    1- Celtic has a full 20 base counting system (including 70,80 etc.)
    2- Danish too
    3- French has partially retained such a system

    From where ? There are 3 hypothesis

    1 - From Scandinavian through the Normans ?
    2 - From the Gaulish substratum ?
    3 - From a pre-IE substratum ?

    I would discard 1 because the influence of Old Norse on French is very poor (no more than 30 words), because Old Norse do not have (apparently) a complete set of 20 base numbers, and because it implies a sudden shift to a more complex system which looks completely unmotivated.

    I would also discard 3, because if it is pre-IE, it should have come to French through Gaulish. Hence it cannot be a direct inheritance

    And I like 2 because it is the most simple explanation and the motivation is obvious (= tradition).

    Understand me : I don't WANT it to be Celtic, I don't care really, I just don't see any other good explanation.
  50. berndf Moderator

    German (Germany)
    By using the word "retained" your argument becomes circular.
    I assume you wrote this before reading #47 and #48. I (and, if I am not terribly mistaken, CapnPrep as well) can't see an advantage of any of the mentioned possibilities over the others (Celtic, Germanic, innovation, other substratum).
    Last edited: Oct 8, 2013

Share This Page