ANATOLY LIBERMAN ON SUPPLETION.

Anatoly Liberman, author of the Oxford Etymologist blog, has a post titled “How come the past of ‘go’ is ‘went?’” He doesn’t actually answer the question, because a real answer is impossible (at one point he says “I consulted numerous books on the history of the Indo-European languages … and discovered to my surprise that all of them enumerate the forms but never go to the beginning of time,” as if it were possible to go to the beginning of time), and he does some fancy rhetorical dancing (“No fully convincing explanation of this phenomenon exists, but some facts can be considered with profit”), but he does provide a handy introduction to the topic with a good set of examples. A warning: though you might think from reading his post that suppletion is an Indo-European thing, it’s not in the least restricted to that family; in Georgian, for example, the verb “to come” has four different roots.

Comments

  1. marie-lucie says:

    The occurrence of suppletive forms is documented in a variety of languages. In most cases suppletive forms occur with words which are very common, and the various forms represent leftovers from a time when the roots of those forms had slightly different though closely related meanings, so that different forms were favoured in slightly different contexts and were later considered synonymous, causing the less frequent usages to disappear. Went is supposed to be the original past tense of to wend, as in the still current to wend one’s way which implies some duration. Perhaps to go (or its ancestor) did not imply duration and therefore was not used in the same contexts. Another example of suppletion is better, best as suppletive forms for good ( similar forms in German and Scandinavian languages show that the suppletion is very old): the root *bet (or something very similar) also meant good, but was probably used mostly, then later only, in the context of comparing degrees of goodness (similar examples exist in Latin, although with completely different roots than the Germanic ones: *bon- versus *mel-). In some other languages suppletion occurs between singular and plural verb roots. As an example, an action such as “running” performed by a single person or by a group or crowd does not make the same impression on a viewer: only in the case of a crowd can one use “pouring” as a synonym to describe the action, and the equivalent of “to pour” can thus become the plural form of a verb with a different root used when describing the action of a single runner.

  2. David Marjanović says:

    My comment isn’t getting through over there, not even when I take the HTML out; there’s no mention of moderation or anything. So I’ll burden you with it:
    ===================

    but I cannot be multiplied, even though grammar says that we is the plural of I. Therefore, it does not come as a surprise that I and we have different roots. Likewise they is not the plural of he, she, or it.

    Uh, it is in Mandarin. I = wǒ, we = wǒmen, he/she/it (when “it” isn’t just zero) = tā, they = tāmen. The same holds for “you”, both normal and especially polite. The suffix -men is almost never applied to nouns, but it’s still obviously a suffix!

    The merger of synonyms within one paradigm may not have been the only source of suppletion, but it was an important one.

    That’s what happened to the Romance “go” words. Their forms are cobbled together from three Latin verbs: ambulare “go for a walk”, vadere “go away, cede, buzz off”, and ire, good old “go”. The fad for dysphemisms at the beginning of Romance must have had a lot to do with this.
    Speaking of ire, I bet that’s what eode is related to. The Old English dialect of Wessex regularly has eo where the rest of the Germanic world has i (seolfor – “silver”, German Silber).

  3. David Marjanović says:

    to wend

    German wenden “to turn around”, wandeln “to walk around in a very poetic manner”, where -l- once marked frequentatives, correctly implying duration in this case.

  4. David Marjanović says:

    Polish spam! What happened to Turkish spam?

  5. Isn’t the verb to be in French conjugated in the present tense an example of suppletion: Je suis, Nous sommes, Ils sont, but Tu es, Il est, Vous êtes (origination from essere and estar, if my Spanish/Italian are still fresh in my mind)?

  6. The suffix -men is almost never applied to nouns, but it’s still obviously a suffix!
    Technical nitpick: that doesn’t necessarily mean that “I is multiplied”. I don’t know where /men/ came from, but the pluralizing suffixes in Japanese (/-tati/, /-ra/, /-domo/, /-gata/) all etymologically mean something like “(X) and others”, “those with (X)”, “those near (X)”, etc. So arguably something like “watasitati” isn’t a plural pronoun in Japanese, but rather a singular-pronoun-plus-a-suffix. (There’s never a “multiple I” situation, only “I+” situations.) And those suffixes can be applied to things other than pronouns too: senseigata “teachers”, aitura “those guys”, etc.
    On the other hand, we do have /wareware/ “we”, which is derived from /ware/ “me” through reduplication, which usually means straight-up “lots of X” in Japanese. That would seem to undermine his “individuals should not be multiplied” theorem a bit, but I suppose you could cling to the hope that the meaning of the reduplicated /ware/ is closer to “self (the concept)” than “I (an individual)”. I’m not sure whether that’s plausible or not.

  7. Also, re this…
    The main work on the origin of suppletion is a “famous” book written more than a hundred years ago, and it had important predecessors. “Everybody,” as various authors say, knows it. Well, apparently, the book’s fame is not universal, and one can devote long years to the study of historical linguistics and stay outside the group defined by the cover term “everybody.”
    … It’s a bit rich for Liberman to complain about this sort of obscurity and clubbiness, but then fail to include the title of the book in his own write-up. Talk about pulling the ladder up after you!

  8. m-l, on another subject, I just read in the NY Times:

    “By choosing Bergoglio we chose someone who was not in the Curia system, because of his mission and his ministry,” said Cardinal André Vingt-Trois, the archbishop of Paris, in a news conference.

    How does one end up with a name like André Vingt-Trois? It wouldn’t happen in English (except in The Prisoner). According to Wikipedia, His surname, which translates to “twenty-three” in English, would seem to be explained, according to Vingt-Trois, by an ancestor who, as a child or baby, was abandoned and found on the 23rd day of the month. – but that doesn’t explain anything, really. (Feel free to ignore this if it seems too silly.)

  9. but that doesn’t explain anything, really
    (Doesn’t explain why anyone would have chosen or accepted 23 as a surname. I feel fairly sure that it wouldn’t have happened in England, so what made it acceptable in France?)

  10. Stephen C. Carlson says:

    Yeah, I’m disappoint that Liberman didn’t cite the the century-old book known to the priesthood of philology. I wonder if it’s Hermann Osthoff, Vom Suppletivwesen der indogermanischen Sprachen (Heidelburg: J. Hörning, 1899).

  11. Bill Walderman says:

    The forms of the present indicative tense of the French verb être, “to be“, look suppletive but aren’t from an etymological perspective.
    The French forms are derived from Latin esse: sum, es, est, sumus, estis, sunt, but ultimately from a Proto-Indo-European athematic verbal root reconstructed as *Hés- (sing.) alternating with *Hs- (plur.), where H is a “laryngeal”. The reconstructed forms are supposed to have been *Hés-mi, *Hés-i, *Hés-ti, *Hs-més, *Hs-tHé, *Hs-énti.
    But the French passé simple of être — je fus, tu fus, il fut, nous fûmes, vous fûtes, ils furent — is indeed suppletive (from Latin fui, fuisti, fuit, fumus, fuistis, fuerunt).
    So are the present indicative forms of the French verb aller, “to go”: je vais, tu vais, il va, nous allons, vous allez, ils vont — which are derived from two different proto-French verbs.

  12. @AJP: There’s a New York Times reporter named Jennifer 8 Lee, here’s an item with her byline:
    http://bits.blogs.nytimes.com/2013/02/27/sergey-brin-explains-google-glasses/?ref=jennifer8lee

  13. Trond Engen says:

    This may be a good time to tout Piotr Gąsiorowski’s blog Language Evolution. His most recent posts are about the suppletive stems of to be. (Earlier he did a series on the number 1.)

  14. It’s a bit rich for Liberman to complain about this sort of obscurity and clubbiness, but then fail to include the title of the book in his own write-up.
    Thanks for saying that; I’d meant to mention that annoying fact in my post but forgot.

  15. This may be a good time to tout Piotr Gąsiorowski’s blog Language Evolution
    Thanks, I’ve added it to my Google Reader feed! (Though how I’m going to keep up with blogs once goddam Google kills off GR in a few months, I don’t yet know. I hear good things about The Old Reader, and will probably give it a try as the fateful day draws nigher.)

  16. Thanks, Matt. I think first names are often more idiosyncratic than last names, though.

  17. mollymooly says:

    Some suppletions are incomplete, contested, or in progress:
    few>less>least
    person>people
    go [to hell and back]>went [to hell and back]>have been [to hell and back]
    wreak [havoc]>wrought [havoc]
    There’s also the odd case of French “jeune homme”>”jeunes gens”.

  18. I’d have thought that “By choosing Bergoglio we chose someone who was not in the Curia system, because of his mission and his ministry,” said Cardinal André Vingt-Trois, the archbishop of Paris, in a news conference” is remarkable for its ambiguity. Praps it’s clearer in French. Or Latin.

  19. mollymooly says:

    Osthoff’s book seems to be online here.

  20. Really? I’m trying to find the ambiguity, and I can’t seem to. Where are you seeing it?

  21. Osthoff’s book seems to be online here.
    God bless the internet! It was fun to search the stacks of Sterling for such items, blow the dust off, and sit cross-legged in the aisle studying them, but it sure is more convenient to sit in my own study and read them at my leisure.

  22. Le Monde blog:

    il n’est pas du système curial, qu’il n’est pas du système italien mais que par sa culture et son enracinement [le nouveau pape a des origines piémontaises], il est italo-compatible

  23. marie-lucie says:

    AJP: André Vingt-Trois
    This name is indeed very unusual. The proposed origin of the name may reflect the practice of an orphanage, a few centuries ago, to use the admission number of an abandoned child of unknown parentage as a a substitute for a last name (there may be other known instances of such a practice, justifying the interpretation). Such a child would have been given a “Christian name” on baptism, but had no last name as he had no known family (the child of an unwed mother would have used his mother’s first name as his last name*). If there were two or more children with the same first name, their numbers could have been used to differentiate them. Once out of the orphanage and needing a last name, the ancestor of the current archbishop may later have kept this number as a last name, not having a right to another. Alternately, it could have been a nickname given to or adopted by the ancestor because of some incident in which the number 23 was relevant, but this is perhaps less likely.
    *In Les Misérables, the hero, Jean Valjean, an escaped convict, later reappears in society as “Monsieur Madeleine”. For a long time I wondered why he had chosen a woman’s first name as his new last name. I think that Madeleine was probably the first name of his mother. The female name, a very common one in France, implied a very humble origin and made it impossible for the police to trace birth records or search for relatives.

  24. I’m surprised that he makes this claim:
    > But in no language are the words for “girl” and “woman” derived from those for “boy” and “man.”
    which is easily shown to be false — consider Hebrew, for example — and does not seem essential to his point.
    (Unless by “no language” he means “no Indo-European language”, which would accord with the complaint in your last sentence.)

  25. Thanks, you’ve confirmed that it’s as odd in France as elsewhere. 23 is a prime number, but I can’t see any significance, unfortunately. About Jean Valjean, are you saying that almost any surname that is also a female first name would imply humble origins? Surely no stigma, though, nowadays?
    dearie, Francis has been a Jesuit priest since he was young. Jesuits typically work in poor neighbourhoods, in poverty themselves, a long distance from the Curia fat cats and any associated corruption. I think it’s great that they chose a Jesuit, not that anyone cares what I think.

  26. I agree, for what it’s worth, which is next to nothing.

  27. J.W. Brewer says:

    I suppose in ecclesiastical circles the last name “Vingt-Trois” might be taken to suggest an affinity for Pope John XXIII, but if you trust wikipedia the Archbishop in question was so named from birth (as was his father before him), before the Pope in question had acquired that name and number. I don’t know how easy or hard it is to switch surnames as a cultural or legal matter in France, but it’s reasonably common in the U.S. and I would think that odd-sounding surnames best explained by an ancestor having been an abandoned foundling who was taken in and renamed by bureaucratically-minded nuns would tend not to be preserved over the generations.
    I don’t know to what extent in French onomastic culture many/most/all first names can be trnasformed into some not-too-weird-sounding surname that’s originally a patronym — but if so, if I had been running a religious foundling hospital I might have assigned surnames derived from the name of whatever saint happened to be commemorated on the day the particular baby had turned up on the doorstep, as being both auspicious and less likely to be a subsequent handicap than a number.

  28. dearieme says:

    @Hat: “By choosing Bergoglio we chose someone who was not in the Curia system, because of his mission and his ministry,”
    Does that mean that they chose him because of his mission and his ministry, or that he was not in the Curia system because of his mission and his ministry, or that they chose him because he was not in the Curia system? These are logically distinct possibilities.
    Anyway, he’s a rash choice since he is now going to be endlessly accused of taking one side or the other in the Argentinian civil war between two bunches of murderous thugs.

  29. Does that mean that they chose him because of his mission and his ministry, or that he was not in the Curia system because of his mission and his ministry, or that they chose him because he was not in the Curia system? These are logically distinct possibilities.
    Ah, I see what you mean. I assumed it meant they chose him because of his mission and his ministry, and (as an additional point of information) he is also someone who was not in the Curia system, the relation of the Curia bit to his choosing being undefined. Two factors made other possibilities invisible to me: 1) semantics (my reading seemed the only sensible one given what I knew about the church and the process), and 2) punctuation (if it had meant he was not in the Curia system because of his mission and his ministry, the comma would not be there, in my scheme of things).

  30. David: dysphemism (had to look up the word, I confess) has nothing to do, I think, with suppletion involving the verb “to go” in Italo-Western Romance.
    IRE was originally a mildly regular verb (eo, is, it, imus, itis, eunt…), but as a consequence of sound changes in the transition from Latin to early Romance many of the shorter forms were simply too ambiguous (thus, EO must have become homophonous with the reflex of Classical EGO in much of the Empire). As a result VADERE was used, especially where the stem /i/- was followed by a non-syllabic ending. Hence the Old Spanish present tense of “to go”: vado/vo, vas, va, imos, ides, van.
    Even this two-stem system, however, was not satisfactory over time: forms of the i-mos, i-des type were very liable to be (mis)interpreted as thematic vowel + ending instead of stem + ending.
    In Spanish this problem was solved by extending the range of VADERE wherever a “zero morpheme” was present: hence Modern Spanish vamos, vais which replaced imos, ides. In the Central part of the Empire (France and North-Central Italy), however, stems deriving from a third verb, ANDARE in Italy, and *ALLARE in Gaul (whose etymology is unclear), were used to fill these “gaps”: hence French allons, allez and Italian andiamo, andate.
    Romanian, in its splendid isolation, came up with a much more straightforward solution to the problem and wholly dumped IRE and replaced it with a single new verb, A MERGE, whose original meaning was “to dive”.

  31. The “because of his mission and ministry” line doesn’t appear to be in the original french.

  32. Surely Madeleine as a surname for an escaped criminal has something to do with the Magdalene Asylums, “institutions from the 18th to the late-20th centuries ostensibly to house ‘fallen women’” (Wikipedia). As a last name it would imply that the wearer was born out of wedlock in such an establishment, that his father was very likely unknown, and that it would therefore be “impossible for the police to trace birth records or search for relatives”, so they shouldn’t bother to try.

  33. Hat, and any other Google Reader users:
    The first thing to do is that as long as you are still using Google Reader, you must use Google Takeout to download your subscriptions file every time you add or remove a subscription. That way, you won’t lose anything when Google Reader finally vanishes. Click on the “Create Archive” button and then the “Download” button, and you’ll be asked for your password and then you’ll get a zip file. Open this file with any zipfile reader and extract the “subscriptions.xml” file. Almost any RSS reader will accept this file and allow you to re-create your subscriptions from it.
    Armed with this file, I looked at The Old Reader, but I haven’t been able to get it to fully accept my subscriptions.xml file just yet. Indeed, I get the rather dispiriting message “There are 28341 users in the import queue ahead of you.” So I turned to another reader, Feedly, which does not require a subscriptions.xml file though it will accept it; you can just tell it to download your current situation directly from Google Reader, and use both of them concurrently. The only disadvantage to Feedly is that you have to download a browser plugin to use it, and it is limited to Firefox and Chrome only: you can’t read your feeds from a friend’s Windows box that only has IE on it. Feedly front-ends Google Reader itself, so it knows not only your subscriptions, but also what posts you have and haven’t read already. When Google Reader goes away, Feedly will be running its own service, or so they are promising now. In any case, make sure your subscriptions.xml file is up to date.
    To make Feedly a little more like Google Reader, go into “Preferences” and choose “Full Articles” as the default view. The “Today” feed contains not today’s articles but your unread articles; the “All” feed displays just unread articles if there are any, but if none, all articles are shown. The check-mark button marks all displayed articles as read (which you have to do by hand in the “Today” feed).

  34. Thanks, sounds like Feedly is worth trying. But knowing me, I’m not going to do a thing until Zero Hour is a lot closer.

  35. I wanted to note that in the Na-Dene languages there’s suppletion in verbs of motion and verbs of handling, but rather than being tense-associated it’s instead plural-associated. So Tlingit has át x̱waagút ‘I went there’ using the root √gut ‘sg. go by foot’ and át wutuwa.át ‘we went there’ using the root √.at ‘pl. go by foot’. Compare this to x̱waax̱áa ‘I ate it’ and wutuwax̱áa ‘we ate it’. The verbs of handling also supplete for different categories of object, so that objects are classified partly by the verb root with which they’re handled. So Tlingit has xʼúxʼ át aawatée ‘he put a book (generic) there’, kʼwátʼ át akaawatée ‘he put an egg (small round) there’, and kées át akawsitée ‘he put a bracelet (framelike) there’, but gúxʼaa át aawatán ‘he put a cup (empty container) there’, gúxʼaa át aawa.ín ‘he put a cup of something (full container) there’, kʼoodásʼ át aawa.áx̱ ‘he put a shirt (flat flexible) there’, kʼoodásʼ át awlináa ‘he put shirts (flat flexible bundle) there’, x̱áat át awsitáa ‘he put a fish (dead animal) there’, káaxʼ át awsinúk ‘he put a chicken (live animal) there’. There’s lots of suppletion in the Na-Dene family, but it’s not in areas of the language that Indo-European linguists are accustomed to.

  36. bruessel says:

    I listened to the news conference of the French cardinals https://soundcloud.com/cathobordeaux02-1/conf-presse-cardinaux-francais and Cardinal Vingt-Trois said: “il n’était pas dans le système curial par sa mission et par son ministère”. So, as there was no comma in the original, to me that means “he was not in the Curia system because of his mission and his ministry”.

  37. marie-lucie says:

    JWB: I don’t know how easy or hard it is to switch surnames as a cultural or legal matter in France, but it’s reasonably common in the U.S.
    I think the US is unusual among countries in this respect, because early immigrants embarking on a “new life” were encouraged to break ties with the old one, including picking new names. The few instances I have personally heard of in France involved immigrants with long foreign-sounding names, who were allowed to shorten or adapt them as part of the naturalization procedure (eg “Mogilewski” became “Moguy”). Native-born French people rarely try to change their names, although this is allowed under certain conditions (including having a ridiculous-sounding name) but the procedure is long and cumbersome. Since women as well as men legally keep their birth names all their lives (although usually being known socially by their husbands’ last names), they don’t have to petition a court to change their names in case of divorce, as happens in North America.
    if I had been running a religious foundling hospital I might have assigned surnames derived from the name of whatever saint happened to be commemorated on the day the particular baby had turned up on the doorstep, as being both auspicious and less likely to be a subsequent handicap than a number.
    The name of the saint of the day would probably have been assigned as a first name (for a long time a tradition especially in rural families) to a newly rescued baby not known to have been baptized. Some of those babies would arrive at the foundling hospital with some piece of identification (such as a paper with a first name). But the occurrence of many male names as last names may also have evolved from using the day’s saint as a last name for an unknown child.
    In French the equivalent of “last name” is nom de famille, the family name usually transmitted in the father’s line. For centuries, an unknown (often irresponsible) father meant no family name for the child. I suppose that adults brought up in those circumstances would choose a last name when they needed to produce one, often a nickname that already identified them, or some other word with some meaning to them. For instance, names like “Lajeunesse” (‘youth’), “Laframboise” (‘raspberry’) or “Laliberté” (‘freedom’) often appear in older literature as names of servants or other subordinates, and such names now current as family names most likely started as assigned nicknames or individually chosen names. That’s what I think might have happened with “Vingt-Trois”, whatever its origin.
    AJP: are you saying that almost any surname that is also a female first name would imply humble origins? Surely no stigma, though, nowadays?
    A child born out of wedlock, unrecognized by its father, and recorded only by its mother’s first name instead of a family name, would ipso facto have been considered as of humble origins in the period described in Les Misérables. The law no longer denies a child a last name, and if there is no recorded father the child’s last name is that of its mother. I am not aware that there is currently a stigma about female first names used as last names, but there would have been at the time of Monsieur Madeleine. These last names are relatively rare, except in the French Caribbean islands because of naming practices sometimes dating from the days of slavery.

  38. marie-lucie says:

    MHendery: Madeleine as a surname for an escaped criminal has something to do with the Magdalene Asylums, “institutions from the 18th to the late-20th centuries ostensibly to house ‘fallen women’” (Wikipedia).
    That would be a possibility in an English novel, but I don’t recall reading about the name being used in French for such establishments. Wikipedia.fr gives the word as denoting (among other things) some kind of hospital, hospice or shelter, originally for travellers, but does not even mention women. The most fashionable church in Paris is called La Madeleine, probably because it was built on or near the site of such a hospice, but it is unlikely that it would have retained a name associated in the public mind with “fallen women”.

  39. marie-lucie says:

    James C., thanks for the Tlingit examples. My examples (given in English but not about English, for singular/plural suppletion) come from not too far from yours.

  40. marie-lucie says:

    comments about the new pope
    There are two sets of comments from French sources: an official one from the archbishop of Paris, contrasting his “mission” and “ministry” to the shenanigans in the Roman Curia, the other one from a blog, about his Argentine nationality and Italian roots allowing him to bridge the cultural gap between South America and Italy.

  41. marie-lucie says:

    La Madeleine (the Paris church)
    I guessed wrong. Wikipedia gives the history of the church in great detail, and it has nothing to do with a supposed hospice, etc, let alone one for “fallen women”.

  42. For reasons best known to himself, my grandson’s father never got around to attaching his name to the birth certificate during the legal period for doing so, and consequently my daughter’s son is surnamed Cowan. I had long since reconciled myself to not having descendants named Cowan, and now I most likely will.

  43. rootlesscosmo says:

    @marie-lucie: they don’t have to petition a court to change their names in case of divorce, as happens in North America. No longer true in California. You can be known by whatever last name you like–with government agencies, banks, whatever–provided you aren’t changing your name in order to commit a fraud. It can be tedious for women to get driver’s licences and insurance etc. changed post-divorce, but they don’t need the court’s authority to do it.

  44. Indeed, that’s true everywhere in the U.S. and always has been, except for a while in Hawaii. Courts do not consummate name changes; they merely take judicial notice that the change has already occurred. When Mr. Kabotchnik of Boston changed his name to Cabot, the famous Cabot family went to court to try and stop him, but were non-suited; as a result, the Official Boston Toast was changed:
    Well, here’s to good old Boston,
    The home of the bean and the cod,
    Where the Lowells talk only to Cabots,
    And the Cabots speak Yiddish, by God.

  45. Trond Engen says:

    I’m always wrong when I try to say something about Icelandic, but as I understand it, the Icelandic system would in principle assign metronymics when the father was unknown, but more often the mother’s father’s name was used as a faux-patronymic to avoid the stigma. With the advent of feminism metronymics became more popular even for children of properly coupled parents, removing the stigma. Or maybe rather making that the least stigmatizing option.

  46. Trond Engen says:

    In Medieval times there was a little church consecrated to Saint Mary Magdalene in a hamlet along the road through the forest some 35 km southeast of Trondheim. That’s a rare thing for Scandinavia, and it’s even stranger since it’s not associated with any sort of hospital. I think the explanation is pragmatic. The feast of Saint Mary Magdalene is July 22, one week before the great feast of Saint Olaf was celebrated in the city, attracting pilgrims and other travellers from near and far. They would have been passing through the hamlet in great numbers around that day, and a convenient local feast would have encouraged them to stay for a night or two.

  47. marie-lucie says:

    Details may vary among different states and Canadian provinces about women changing their names after divorce, but it is usually something official. When I first came to Canada I was surprised at the number of notices in newspapers placed by women with very ordinary names changing them to names that were just as ordinary, but no name changes for men. I found out later these were divorced women regaining their “maiden” names, not changing them to names they liked better on esthetic grounds.

Speak Your Mind

*