1. mldh2o New Member


    I was just investigating the word mother, which seems to be similar in an awful lot of languages.

    English: mother (Old English: modor)
    German: Mutter
    Dutch: moeder
    Swedish: moder
    Yiddish: muter
    Icelandic/Faeroeese: móðir

    Czech/Polish: matka
    Slovak: majka
    Russian: мать (mat')

    Gaelic: mathair

    French: mère (does anyone know the Old French?)
    Spanish/Italian: madre
    Portuguese: mãe

    Chinese: mŭqīn

    Latin: mater, matris
    Ancient Greek: μήτηρ (méter)
    Sanskrit: matri ???

    Does anyone have any contributions to this list, especially from outside europe? Why should this word be similar in so many languages when other words don't remain so close? Is there a good reason why the romance languages seem to drop the dental sound from the middle (hence why I'd like to know the Old French or even Vulgar Latin).

  2. Mahaodeh Senior Member

    Arabic and English
    I can add Arabic = umm = أم
  3. mldh2o New Member

    How would you pronounce that? With the two ms separately or together?
  4. Mahaodeh Senior Member

    Arabic and English
    The double m is to show that there is a stress on the m, not that there are two letters.

    The interesting thing is that mother is umm, father is abb (also a stress not two letters). children call their mothers "mama" and fathers "baba". It sounds so much like the French way of calling parents (mama and papa); I doubt that the words are related though. It's probably a conciendce.
  5. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    As for 'father' and 'mother' in Indoeuropean languages (this includes most European languages - except Basque and some other languages not suiting into any group, spoken especially in the Caucasus, and the Finno-Ugric and Turkish-Mongolic language groups - and Sanskrit-related Indian languages as well as Persian and Afghani): in Indoeuropean languages, the words of 'mother' and 'father' go back to the same stem, so they all look very similar.

    There do exist even theories about similarities with the Semitian/Hamite language stream, meaning there might be a substrate Pre-Indoeuropean and Pre-Semite where 'father' and 'mother' might come from:
    But me, I think that it's a bit unlikely that 'umm' and 'abb' go back to the same stem as the Indoeuropean - and anyway, it would be extremely hard to prove the connection there.

    As for 'mama' and 'baba', this is a different thing: both words consist of the most simplistic syllables which are among the first most children are able to speak:
    - labial nasal + A = BA- (PA-)
    - labial plosive + A = MA-
    So it seems quite logical that kids, speaking their first words with having in mind to put a name to the most important people in their young lives, will use these simple syllables
    (dental plosives and sounds are more difficult to utter for children at that age as this requires more complex moves with the tongue, wheras with labials they only have to close and open the mouth - and when open, A is the easiest vowel to utter)

    [EDIT: Well, just one afterthougt: it is strange that in this child's talk, 'MA' always stands for the mother and 'BA/PA' always for the father; there's a similar thing in Somali, I think, but it could be native there similarly to Arabian; probably someone speaking African or Non-Indoeuropean Asian languages may contribute to this child's naming of parents; I wouldn't trust anyone from Indio background in Southern America as there, most likely, the European influence should have been overwhelming.]
  6. mldh2o New Member

    Do you think there's any relation between the Chinese muqin and the Indoeuropean root, Chinese not being directly from this? I notice the Japanese okasan bares no resemblance.
    It does seem like the q in the middle of muqin (a velar) is a long way from the t and d that most Indoeuropean words have (dentals), but from a Chinese pronunciation it might not be so far flung. I know Chinese characters developed rather differently to Roman ones, so a slide in value might be plausible.
  7. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    It's very, very unlikely - though not impossible; if there is something, anything to the theory that a few of the most basic words on the world (including, of course, 'father' and, even more so, 'mother') have roots in a common 'human' stem of words (the theory goes so far as to suggest that the first humans had a small set of common words - say, 100 or a little bit more), then it might be possible.

    But it would be hell to prove something like that. ;-)

    Easier might be to prove a connection between two or three language stems, say between Indoeuropean and Hamito-Semitic stem, and even Indoeuropean and Mongolic stem, as all of them had a geographical connection - but Chinese is not exactly part of the Mongolic stem ... so the brigde you'd have to build would have to be even longer.
  8. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    All the languages above belong to the same family (Indo-European), and I'm pretty sure that all of their words for "mother" come from the same Proto-IE root. (Interestingly, in Albanian, this word drifted semantically and came to mean "sister".)

    For Indo-European languages, the answer is that they all descended from the same pre-historical ancestor language, and many of them retained this word in a somewhat similar form.

    However, you are right that the words for "mother" show more than accidental similarity even across languages that are (to the best present linguistic knowledge) unrelated. There is an excellent article that presents a convincing explanation for this phenomenon here:

    To summarize its points briefly, the explanation is that the first babbling noises that small children start making usually sound like ma-ma or something similar. The reason is straightforward: these are the simplest sounds to produce for a child whose muscles of tongue and mouth haven't yet developed fully. Just keep opening and closing your mouth with your lips and tongue relaxed, while voicing your vocal chords, and you'll produce babbling sounding vaguely like ma-ma-ma... Of course, the mother is apt to interpret this meaningless babbling as the child addressing her, which sometimes results in her and other members of the household picking up a word similar to the child's babbling as their word for "mother", and occasionally such words spread to the whole language. My summary doesn't really do justice to the article (I've oversimplified or ignored many important points), so I'd definitely recommend reading the article if you're interested in the topic.
  9. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    Thank you, Athaulf, the article which you've linked above is exactly about what I wrote (only I was to lazy to dig up something like that ...).

    As for the theory, seems it mostly was postulated by Roman Jakobson although this paper on the theory from the website of the University of Sussex seems to be unsigned. The core of what's said in the paper is:

    - the most simple syllables a baby will utter will be PA and MA or something like that
    - parents tend to believe that babies will refer to them with the first sounds they utter, and as these usually would be PA and MA it is only logical that they would become attributed to the meanings 'mother' and 'father' (this I described in my posting above the other way round, but I think Jakobson's explanation is much more accurate ;-)

    So, no big mystery as to why 'mother' and 'father' everywhere in the world sound rather similar. (If you allow, of course, for language change taking place over the generations so that you have different forms of the word for 'mother' in different Indoeuropean languages and alternatively, too, the 'original' baby talk 'mama' or similar: in this case, it would be logical for parents to compare MA with 'mother" and so attribute to baby talk 'mama' the meaning of mother.)

    But - this does not answer another question which arises if you read the paper (page 2):
    - in 52% of the languages observed by Murdoch, the syllable MA, ME or MO was an element in the word for 'mother', and only in 15% in the word for 'father', and:
    - in 55% of the languages the syllable PA or PO, TA or TO was an element in the word for 'father', but only in 7% in the word for 'mother'

    This means: there is no explanation why 'mother' in most languages contains the nasal M, whereas 'father' mostly contains the plosive P (or T), and interestingly only few 'mother'-words contain P/T and even less 'father'-words contain M.

    On the other hand, one would have not to count languages but language groups: it is, of course, logical that in Indoeuropean languages the distribution would be M for mother and P for father (don't forget: English F here still derives from P, and in English you've got the baby talk 'pa(pa)'). So, Indoeuropean should count as only one language - as should the Semito-Hamitic group (if the distribution there is equally consistent, which I do not know).

    One would have to start from scratch to try and prove that indeed there was something like a "Proto-World" language as referred to in the article: but even then you'd have to explain away the exceptions.
    Difficult, if not impossible.
    Anyway, if there were a distribution (counting language groups, of course!) of about around 50% 'M = mother' and 50% 'P/T = father', then this of course would be a very strong argument against such a Proto-World. 60%:40% still would be rather good.
    But if, on the other hand, one would find that the relation would be more like 90% 'M = mother' and 10% 'P/T = father', then this would suggest that probably there really was a Proto-World with a very basic set of words which survived in most (if not all) modern languages.

    But, as the author of the paper says, it would be very difficult (even if there was a Proto-World) to prove it, as languages change: with the English word for 'father' you already have an example for that - but Indoeuropean language changes are well documented by ancient languages as Latin and Greek as well as fragments of lots of other now dead Indoeuropean languages.
    This situation is much worse with most (if not all) other language branches. Even in the case of Chinese, even though there's ancient script too: that's because Chinese is not written phonetically or with syllables but rather a Chinese sign stands for a certain meaning, and phonetical change is not visible in writing.
    (As is, by the way, the phonetical change of English since Shakespeare, but that is another story altogehter.)

    On the other hand, the author of the paper overdramatizises a little bit: yes, there were huge changes in the Indoeuropean languages in the last 6.000 years, but there were also huge socioeconomic changes in this period, beginning with the Neolithic Revolution.
    Before that, the human society and culture was extremely stable and conservative, so one could postulate that language also would have been much more stable than after the Neolithic Revolution. It is absolutely possible for a hypothetical Pre-Neolithic World Language to stay rather stable over a period of tens of thousands of years - it is a hypothesis one could work on. But again: would be hell to prove. Ever.

    So, most likely, we can only put probabilities to any theory postulating a common origin of basic words like 'mother' and 'father'. And to give such a theory a little bit more than minimal probability, one would have to find a little bit more words than just two. This would have to be words of a pre-neolithic society, of course - and no words meaning plants or animals, as these change according to region and climate.

    Really, it is an impossible task ... but nice to speculate about.

    Well, probably I should add one argument, if you read the text of University of Sussex linked above by Athaulf:

    The article describes several cases where the original words for 'father' and 'mother' were replaced by newer baby talk words derived from 'mama' and 'papa' or similar.
    But this is no argument for the distribution of M/mother vs. P/father: because one can safely assume that if such a replacement were to take place, then at least for a longer period of time 'old word' and 'baby talk word' coexist. It would be totally consistent, then, to attribute the 'mama'-word to 'mother' or similar and vice versa (so, to attribute nasal to nasal, and plosive to plosive).

    So, again, crucial would be the 'real' distribution of nasal and plosive for both meanings, in as many language groups around the world as possible. (Sorry, this don't makes the task easier.)
  10. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    Actually, there is an explanation -- and a quite plausible one, I'd say -- on page 17. The easiest, and thus normally the first consonants for small children to produce are the labial nasal [m] and the labial stops [p] and , in that order (none of these requires any effort from the tongue, and [m] almost none from the lips either). These are followed by the somewhat harder dental nasals and stops ([n], [t], etc.), which are still much easier than most other consonants.

    Now, when the children's babbling noises are (mis)interpreted by the adults around the child, we can expect that in most cases, the child's first utterances will be interpreted as referring to its family members roughly in the order of their closeness to the child. Assuming that the mother is the closest to the child (which is nearly always the case), then the very first utterances, normally with [m], will be "claimed" by the mother, and the subsequent ones by the other close relatives (but not as close as the mother) -- which usually means the father, and then the grandparents, the older siblings, etc. Thus words with [m] usually get to mean "mother", and those with [p], , [t], etc. usually get to mean "father" and (less frequently) other close relatives.

    This mechanism, of course, works only with a certain probability, not in 100% of all cases. Sometimes the utterances get interpreted in a different order, but according to the numbers presented in the paper, this happens in a minority of cases. This theory also makes some additional interesting true predictions, for example that the babbling-like words are often used for older siblings, but almost never for younger ones (since the child is normally no longer at the babbling stage when these appear).

    Not necessarily. It could also mean that the above described mechanism works with an even greater probability. Besides, there is lots of evidence that in most languages with a long recorded history or with reliable reconstructions reaching far back, the ma/pa words have actually changed, disappeared, and reappeared, rather than being examples of great stability (see Section 5 of the paper for the details).

    Well, the problem is that a hypothesis like this is that it's, as you say, inherently unverifiable. However, there has been lots of research on the languages of societies that had barely changed technologically for thousands of years until very recently or even until nowadays. I don't think these have shown any great linguistic stability (even without any written records, it's often possible to reconstruct the changes that several related languages have gone through since splitting off from their common ancestor).

    This is an interesting argument. I guess we could observe the known cases of such replacement, and see how much correlation the new words arising from the interpretation of babbling show with the previous words for parents in each language. However, I don't really see how we could untangle the correlation arising purely from the similarity to the existing words and the correlation arising from the above described natural mechanism of mothers claiming the [m]-words first.

    Perhaps we could observe the cases of ma/pa words arising in languages in which the existing words for "mother" and "father" have already passed through sound changes that mutated [m], [p], etc. into something completely different. If the new words arising without the "model" [m]/[p] words still show the same tendency towards [m]-words for "mother", this would effectively disprove your theory. However, I'm not sure if there would be enough data available for such a study.
  11. J.F. de TROYES Senior Member


    Here are some more examples borrowed from languages that don't belong to the same families :

    MOTHER Burmese ( Sino-Tibetan ) ?ame ; Thai ( Tai-kadai ) mâe ;
    Vietnamese me ; Cambodian ( Mon-Khmer ) m'daay ; Kikuyu (Bantu ) maitu ; Kechua (Inca) mama ; but Guarani ( Tupi ) sy

    FATHER Burmese ?ëphe ; Thai phâw Viet. cha (south) / ; Camb. ëu?pu? ; Kikuyu baba / awa ; Kechua tayta ; Guarani túva

    Forms as "pedre" ( "St Alexis", 11th century ) , "pere" ( Chanson de Roland", 12th ) are documented and it's the same for "madre" (9th) changing to "mere" (10th or 11th). To-day "mair/ maire "are used in Occitan,"mare" in Catalan. So the intervocalic "t" in LATIN "matrem" was first voiced "d" as in Spanish , but "d" was gradually weakened and the cluster "dr" ended up dropping in French and other Gaulish- influenced languages. More generally weakening and finally often collapsing plosives in intervocalic position are an outstanding feature of the phonetic evolution from Latin to French :See past participles as Latin "ornatus", Italian "ornato", Spanish "ornado" and "orné" in French or with another plosive : Latin "focum" > fogu > fuogu > fuou > French "feu" ( Italian : "fuoco", Spanish: "fuego", Portuguese: "fogo" )

  12. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)

    Yes, I know, I've read it - but I know too that M and P/B really are first spoken by babies practically at the same time; the explanation according to Roman Jakobson doesn't satisfy me, really: this minor difference (really a tiny one, in terms of simplicity) would make a difference if the distribution M=mother vs. P/B=father were about 60:40% (or even, say, 70:30%), but if one were able to prove that the distribution were rather 90:10% in favour of M=mother, then I personally would be almonst convinced that there indeed WAS something like a Proto-World language.

    Even the difference between P/B and D/T (in terms of simplicity) is rather small and you will hear babies utter D/T very early in their lives, N too for that matter.

    If there weren't a clearly visible, provable preference of M (nasal) for 'mother' then of course this would support Roman Jakobsons theory; but if there were, then that's different - although still impossible to prove, then it would at least be quite probable for a Proto-World to ever have existed. (One could only talk in probabilities here, as already mentioned.)

    True, but as mentioned by me above: nevertheless in these languages (quite obviously) the MA syllable was used for MOTHER (or similar), so the 'new' babbling word with nasal replaced the 'old' babbling word with nasal, and similarly with plosives.
    Then of course there would remain at least some irregularities, as much is clear from the paper above. It wouldn't be possible to explain everything away, no matter what theory you'd prefer.

    Not quite.
    Slavic languages, when they surfaced to the world of written letters in the medieval age, were rather conservative, meaning close to the satem-branch of the hypothetical Indoeuropean ancestor, and even more so were the Baltic languages at the time. Which did correspond, in both cases, with rather little changes socioeconomically, although some Slavic tribes did live rather early in contact with German tribes. There seemed to be very little dialectal variation at the time, too.
    Great sociological and political changes (meaning especial political divisions - new frontiers and new states) came to the slavic world beginning in the 10th century, and accordingly the Slavic world fragmented.

    But of course it always would be different to prove some argument if you haven't got written documents to support them.

    Yes, if one would find examples were originally there was (this is a constructed example, of course - like Einstein did, a 'Gedankenexperiment' ;-):
    - MAMA=mother and BABA=father, but where through changes of sounds MA changed into AYI and BA changed into MAWI (to choose something not totally unlikely: if the M of MA were deleted, then the change of BA into MA might well happen; I did also delete the second M and replace it with a sliding vowel and changed both words to a more difficult one, in terms of simplicity, in order to make the following point possible and unambiguous)
    - and where, further, new babbling sounds of MAMA and BABA arose and these were NOT attributed as MAMA=MAWI (meaning: father) and BABA=AYA (meaning: mother; not attributed because of similarity, but because there is more similarity between MAMA and MAWI) but rather MAMA=AYA and BABA=MAWI, then this would indeed be a very strong argument pro Roman Jakobson 'strong' natural approach and would as good as disprove the Proto-World theory

    But if, on the other hand, there would be attributed BABA=AYA and MAMA=MAWI, then still (of course) there would be no prove for a Proto-World, but this language still would count as being in the 'pool' for languages having 'BA=mother' and 'BA/PA=father', meaning as one of the languages supporting the Proto-World theory.

    All in all: it might be possible, some day, to rule out the Proto-World theory almost completely, whereas it would be highly unlikely ever to prove the Proto-World theory. So the odds most certainly are with Roman Jakobson.

    Note: I do not support either - just playing advocatus diaboli.

    J.F. de TROYES: your examples from Asian languages all do fit in almost perfectly into the scheme of 'MA=mother' and 'BA/PA=father'. Nevertheless, this still doesn't tell us for sure if Roman Jakobson 'naturalist' approach or the Proto-World approach would be the most likely one - or still another explanation would fit even better.
    But thanks for the examples!
  13. Arrius

    Arrius Senior Member

    English, UK
    As Athaulf has pointed out the M sound is the easiest consonant for a baby to make, hence its prevalence as a sound in the words for mother. Here are two Bantu words for mother to add to the collection: amai in Cinyanja and mayo in Chibemba.
    By the time the baby gets round to verbally identifying its father, it has a greater range of consonants, but B is usually used for this purpose, for example in Swahili baba (which in English becomes later the baby's name for itself or its peers), but Cinyanja has tata for father, which is an unvoiced version of English dada. Somewhat perversely dada is also French baby talk for English gee-gee (horse).
  14. Outsider Senior Member

    Portuguese (Portugal)
    I just wanted to add that I get the impression, from what I've read about these words in various languages (there were previous threads in the Other Languages forum, but the search function isn't working, so unfortunately I can't link to them), that many languages of southern and southeastern Asia have borrowed "mama" and "papa" from European languages fairly recently. Never underestimate the influence of colonialism and western cultural dominance. I can't say that I've looked into this carefully, but it seems like a plausible hypothesis, when so many of these languages have other, more traditional, and very different words for "mom" and "dad" alongside the ones that resemble the European words.
  15. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    However, such a state of affairs can also be a consequence of the above described natural process. The existing ma/pa words will eventually mutate into something sounding quite different, and in such a situation, new imitations of children's babbling may easily arise as novel hypocoristic terms for parents. This can of course continue in an endless cycle.

    Something like this has actually happened in English. The old Indo-European ma/pa words have mutated into English words mother and father that now contain difficult fricatives and sound highly formal and traditional, while new hypocoristic forms of the ma/pa variety (mom, dad, pa...) have replaced them in the actual communication between parents and children.

    I don't know about these Asian languages that you mention, but this process is certainly a plausible alternative to foreign borrowing. Of course, borrowing is also possible; in many dialects of Bosnia, for example, the word for "father" has been borrowed from Turkish.
  16. Outsider Senior Member

    Portuguese (Portugal)
    Here's the previous thread, and here's a typical example:

    It seems that when you dig deep enough these "m-m-" and "p-p-/d-d-" kinds of words often turn out to be recent loans from European languages.
  17. Athaulf

    Athaulf Senior Member

    Toronto, Canada
    Croatian/Bosnia, Croatia
    At this point, I suppose that we would need some hard data about the actual frequency of m versus p/b/t/d among the first children's utterances and the average time gap between their appearance, assuming a significant one exists at all. Unfortunately, I have no idea where I could find such information, although I'm sure someone must have investigated the issue.

    However, the problem with such theories is that it's easy to find isolated data points that fit them nicely, but I've never seen an argument that would really justify any generalizations along these lines. In fact, I'd say that:

    (1) It's impossible to satisfactorily define what accounts for great "socioeconomic" changes that would be relevant for language change. What exact historical events do we have in mind? One can of course cherry-pick the definitions, and then further stretch them in each particular case so as to agree with what one wants to prove, but this is an obviously flawed approach.

    (2) Even if there existed an universal agreement on what constitutes relevant "socioeconomic" change, for any reasonable definition of such changes, I can easily think of counterexamples in which language change was remarkably slow despite enormous, sometimes cataclysmic social upheavals. One can of course find counterexamples of the other sort too, i.e. languages that changed drastically in a relatively short period without any obvious socioeconomic upheavals taking place.

    In fact, I can easily use the example of Slavic languages to argue against your claim. For the last thousand years, the history of Slavic peoples has been as turbulent, violent, and chaotic as anything ever seen in world history. They've lived under dozens of different states and empires, professing several different religions, some under domestic masters, others under foreign Slavic masters, others yet under Mongols, Turks, Germans, Tatars, Hungarians... They also passed through every possible social and economic up and and down through that time. And yet, Slavic languages have been extremely conservative and stable through that period. Even the mutually remotest Slavic languages, spoken by peoples living thousands of miles from each other for many centuries under different states, still have at least some mutual intelligibility. Any Slav can still get at least the gist of Old Church Slavonic texts by just learning the old Cyrillic or Glagolitic alphabet, whereas Old English texts from the same period might as well be Chinese for a modern English speaker -- and England has been a paragon of social stability and continuity compared with just about any part of Easter Europe. There is absolutely nothing indicating that Common Slavic was any more stable and conservative before different Slavic tribes spread all across Eastern Europe in early medieval times.

    Now I realize that I've gone into the rant mode a bit :D, but still, I think that these examples (and many others) show that the theories connecting the rapidity of language change with historical events don't have any firm support in historical data.

    Yes, that's exactly what I have in mind. If we could find enough languages in which the sound changes had already obscured the relevant consonants in the previous round of the ma/pa words at one point, then it would be extremely interesting to see the statistical distribution of these consonants in the newly emerging round of such words. However, I have no idea how much reliable raw data would be available for such a study.
  18. Mahaodeh Senior Member

    Arabic and English
    Possible, but I wouldn't say "Indo-European", I'd say European; in my opinion it's most possibly French, Spanish or English due to the 18th and 19th century explorations and colonialism. It works especially well when the letter m and b/p already exist in the name in the native language (as in Arabic - the names are similar but not identical).

    But then again, this is a theory I just made up - it could be a good one and it could be stupid.
  19. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    It would be almost impossible to achieve that.

    Anyway, if one would ever want to prove that there was indeed a kind of Proto-World language then the two words 'father' and 'mother' never would suffice: there should be at least a dozen words, or more (as many as possible), to convince any linguist.
    Me, I'm not convinced at all, I'm just discussing the probability and what would have to do to prove something ...

    To prove this, [...]
    I have created a new thread in which Proto-World Language can be discussed.
    Moderator EHL

    I know, I've only learned Slovene properly, nevertheless I can read (with more or less difficulty, and a dictionary at hand) all other Slavic langauges (well, at least most - I haven't yet tried Sorbic, and I am not yet prepared to learn the Glagolitic script, although Cyrillic don't gives me any trouble).

    Nevertheless, there was change, even in the Slavic world, more so - there were huge changes.[...]
    I have also created a new thread on Slavic languages.
    Moderator EHL
  20. aleCcowaN Senior Member

    Castellano - Argentina
    The first time I read about the """astonishing""" similarities of the words meaning mother and father all around the World was when I was about 14 years old and reading "The Mystery of Atlantis" by Charles Berlitz, and this """scientist""", """serious researcher""":rolleyes: used this argument to propose the world wide influence of this civilization. I was those years involved in reading Erich von Däniken books, reading about a hollow planet Earth, and the extraordinary theories of Velikowsky about Jupiter spitting planets onto the Earth and causing biblical catastrophes. I was swallowed by the epistemological hedonism that governed and govern culture since 1970 and I was mixing up the Friday night movie with reality. Being a little more mature, a few years later, I realised that mother, mamma et al, with a single of double "m" matches baby's sounds and mouth movements while addressing his mother and asking her for a nipple to suck. Not surprisingly, I read that father derives from mother by opposition, with a hard consonant replacing the soft m, following in some way the maluma/takete-bouba/kiki hypothesis.

    Then in my country we say indistinctly /ma'ma/ from Spanish or /'mama/ from Kechua, and Spanish /pa'pa/ Kechua /'ata/ or other Indian languages /'tata/, without ado or Atlantic, Nostratic or another language behind.
  21. sokol

    sokol Senior Member

    Vienna, Austria; raised in Upper Austria
    Austrian (as opposed to Australian)
    The Bouba/kiki-effect is new to me, and thanks for mentioning, this really adds to the original question, and I have to say that this explanation satisfies me a little bit more than the natural approach of Roman Jakobson. (Although I have a very high opinion of Roman Jakobson's work, I think that in his natural analogy concerning MA/BA as described above pushes just a little bit too far.)

    It would be very logical for parents to attribute the 'softer', probably even 'rounder' (though it's no rounded sound, properly) 'M' with the breasts of the mother and the 'harder' P/T (B/D) with the father.

    Maybe someone will make field studies concerning this matter - could result in interesting findings.
  22. da1andonlysupa New Member

    Actually, Sanskrit is matar. The form origingally written was matri, which is the instrumental form of the word. And the Greek was mater in Homeric Greek, which is older than Attic. I believe this would give us IE *matr (the a being long and the r being a sonorant).
  23. demalaga Member

    España castellano
    In most languages , the word for "mother" includes a nasal consonant, an "m" or an "n".This could be due to the fact that mothers usually make a sound to soothe crying babies, and the most appropiated sound is a nasal sound.In Spanish the word for any song to soothe babies is "nana".Also this word is used for the person who takes care of a child (nanny in english)
  24. Frank06

    Frank06 Senior Member

    Nederlands / Dutch (Belgium)
    Hi everybody,

    It is really not necessary to give as many translations for "mother" as possible (see this OL thread).

    Thanks to everybody who already contributed and who did concentrate upon the question:


    Moderator EHL
  25. J.F. de TROYES Senior Member

    muqin is a compound made of two words the second of which qin is found in fuqin, but also in qin ai (dear ) or qin qie (friendly ) because qin is a kind of suffix which means "closely related".
    As you see "mother" and "muqin" cannot be etymologically related.
  26. J.F. de TROYES Senior Member

    I do know that in such questions, chiefly when this kind of word is concerned, the origin may be surprising. However French dictionaries come up with another one by assuming that "dada" could come back to " dia", a word carters used to shout to their horse when ordering them to turn left. This "dia" comes from the interjection "da" used with "oui" : "Oui-da" (yes, sure! ) made up itself of the two words "dis+va" ( Say!go! ).
  27. J.F. de TROYES Senior Member

    Thanks too, aleCcowaN , for this information I have already read about, but forgotten. Similar experiments were carried out to answer the question whether a mere shift of phonems could be related to a couple of opposite sensations or notions. Sapir showed in 1929 that both meaningless syllables "mi" vs "ma" matched the opposition "small" vs "big" for American speakers ( and I am pretty sure it would be the same for French speakers ). Unfortunately most of this studies (or all ? ) are based on surveys taking only one or two languages into account.
    However I agree with you to think that these data result in a stronger argument than R.Jakobson's point of view based on the order according to which sounds arise in babbling talks.

    After Jakobson's work several, but fragmented studies have been carried out with Corean, Japanese, Quechua, Swedish, Italian, English, French children. I have only read a summary of these data. On the whole :
    1. There is continuum between the babbling period and the word-uttering period, what could explain slight differences in the order of appearance between various languages ( maybe the dorsal plosives could appear sooner or later )
    2. Chidren start preferably forming open syllables ( Cs+ Vowel ) (90%)with labial plosives ( b, p m ) , coronal /t, d ,n/ (27%) and in a lower extent semi-consonants [w,j]. Only 7% would be dorsal consonants. Nothing is mentioned about how these utterances are spaced out in time and I suppose it's impossible to make sharp differences, especially between labial and coronal ones.
    3. Labial plosives are preferably associated with /a/, coronal with front or near front vowels [di, toe ...], dorsal with .

    So I think Jakobson's explanation of the common origin of mother-meaning words turns out to be strong, but I am not convinced by his view on the opposition M-mother vs P-B/T-D-father.
  28. Erutuon Member

    English, USA
    The reason for the first ē of Attic mētēr is that Attic changes an original ā to ē except before e, i, and r. (Ionic, on the other hand, changes it in all cases, while Doric leaves it in all cases.) This was an open long ē, roughly English bed.
  29. da1andonlysupa New Member


    Yes, I believe the ancient presence of the digamma is the culprit for those exceptions.

Share This Page