Encyclopedia of Attested Semantic Shifts.

Christopher Culver writes about “Some semantic shifts found in motion verbs” (“in Old English, wendan meant ‘to turn’. Thus we find the shift ‘he turned’ > ‘he went’”; “Romanian a merge ‘to go’ < Latin mergere ‘to sink’”) and concludes:

It would be nice if there were some massive encyclopedia of attested semantic shifts in the languages of the world. To defend one semantic shift in Chuvash that I claim in a paper I’m working on, I’ve had to gather cross-linguistic support for such a change either from the trivia that I myself know, or from hitting up my colleagues working with different languages for the trivia they know.

What a great idea! It’s so great, and so obvious once you think of it, that it’s hard to believe there isn’t such a thing. Somebody make it happen!

Comments

  1. Wm Annis says:

    I recently ran across a paper — with support data! — on using clustering software to build networks of cross-linguistic polysemies. Not quite what you’re talking about, but pretty close if we take polysemy to be a semantic shift in media res.

    The paper: http://aclweb.org/anthology//W/W13/W13-0208.pdf
    The data: http://aclweb.org/supplementals/W/W13/W13-0208.Attachment.zip

    The file ‘in_out_edges_per_community.txt’ is the goldmine in a more or less readable format.

  2. You might be interested in a project called the Catalogue of Semantic Shifts, headed by (afaik) Anna Zalizniak. Here’s the paper that describes the catalogue:

    Zalizniak, Anna A., Maria Bulakh, Dmitrij Ganenkov, Ilya Gruntov, Timur Maisak, and Maxim Russo. 2012. ‘The Catalogue of Semantic Shifts as a Database for Lexical Semantic Typology’. Linguistics 50 (3): 633–69.

  3. Trond Engen says:

    if we take polysemy to be a semantic shift in media res

    One more!

  4. I’ve often had this same thought, not just about semantic shifts but other kinds of change as well, e.g. sound change. (I think Claire Bowern was working on a database of known sound changes, but I don’t know if this came to anything.) The project Sigva cites looks interesting, though I can’t tell from skimming the paper whether the database is publicly available. You’d think the best way to do this kind of thing would be as a wiki, where experts on particular languages could openly submit exemplars. Language Change Wiki, anyone?

  5. There are several diagrams of semantic shifts in Australian languages, some surprising, in:
    David Wilkins, Natural tendencies in semantic change and the search for cognates. In The Comparative Method Reviewed : Regularity and Irregularity in Language Change, Mark Durie and Malcolm Ross, eds. 1996. p. 264

  6. Language Change Wiki, anyone?

    Make it so!

  7. Wm Annis says:

    The Zalizniak project does have a web site: http://semshifts.iling-ran.ru/semantic-shifts/show-all-shifts.html

    I also had this brought to my attention recently: Semantic Maps for Sino-Tibetan: http://stedt.berkeley.edu/~stedt-cgi/rootcanal.pl/chapters . Those with checks under the “Flowchart” column have actual maps. Tidier arrangements can be found in this list of PDF documents: http://stedt.berkeley.edu/etymologies

  8. Those links make for fascinating browsing, Wm Annis. Interestingly, most of the “shifts” in the Zalizniak database are actually examples of synchronic polysemy patterns, but that makes sense when you think about it (since most shifts go through a polysemy phase and since the latter are much easier to find and collect).

  9. Stefan Holm says:

    Semantics is a jungle, even if you like Culver just stick to motion verbs. Here’s a brief Swedish lesson about hasty movements, perhaps amusing from an English point of view:

    Löpa (cf. leap) is run, typically on the sport’s track. The nouns löpare is then ‘runner’ or the bishop piece in chess and lopp is ‘race’. Springa is also ‘run’ but referring to children’s or other non professionals’ running. The corresponding noun, språng, is though ‘leap’ (Mao’s great leap is ‘det stora språnget’). Another derived noun, springare, is either a steed or the knight piece in chess. Ränna (cognate to run) means run but mostly derogatory – about a person ‘running around’ (and sticking his/her nose in just everything). Another cognate rinna is however neutral for the running of liquids.

    As for ‘jump’ the verb is hoppa and the noun hopp. But there is also the verb jumpa, which is however restricted to the maybe not so healthy boys’ game of jumping on ice flakes. In a play with words it has though inspired the common (slang) noun for school gymnastics, gympa. ‘Hop’ we usually translate with skutta, unfortunately of unsure etymology.

    So a glimpse from just this limited area makes me fear fear that an encyclopedia of semantic shifts would be a never ending story and a mission impossible to carry out.

  10. Jumping on ice flakes?

  11. Stefan Holm says:

    During winter the lakes in Sweden (and we’ve got some 10.000) are covered with an ice sheet. When it starts to melt and the sheet breaks up into flakes boys for some reason have always had a fancy for jumping on those from one to another to the horror of their parents.

  12. Stefan Holm says:

    Hmm, checked it up and saw, that Swedish ‘isflak’ is ‘ice floe’ in English. There you see, what semantics can do between related languages. (Of course those ‘flakes’ are a little bigger than corn flakes).

  13. löpare is then ‘runner’ or the bishop piece in chess [...] springare is either a steed or the knight piece in chess

    Probably calques of German Läufer ‘chess bishop’ and Springer ‘chess knight’.

    English run the gauntlet ‘run between two lines of soldiers who beat you as you pass’ (a military punishment, now used figuratively for any ordeal) involves a disguised version of Swedish gatlopp ‘passageway’. It was borrowed in the Thirty Years War, but changed in the 18-19C under the influence of gauntlet ‘armored glove’.

    Run is a very strange verb. In Scots it is rin, ran, run, a regular (so to speak) strong verb, cognate with German rinnen ‘flow, run, trickle’ and as such intransitive. In English, though, it merged at some point with the descendant of its own causative *rennjan originally ’cause to run’, later ‘run to, ride to, reach by running or riding’. In standard (that is to say West Saxon) Old English, this underwent regular vowel breaking and irregular metathesis to surface as earnan, but in the West Midlands varieties that led to the modern standard, neither of these things happened.

    When looking at the usual list of sound changes between Old and Middle English, it’s important to realize that some of them, specifically including the undoing of vowel breaking, didn’t really happen. It was Wessex Old English that innovated here before being written down, and London Middle English that retained the primitive sound, the Old English change being lost (at least to the standard). This accounts for the apparent violations of Dollo’s Law.

  14. Oh, I see. Nothing like snowflakes, then. By the way, apparently ice floes were called flakes by English-speaking arctic explorers until about 200 years ago.

  15. By the way, apparently ice floes were called flakes by English-speaking arctic explorers until about 200 years ago.

    I’ll be damned — the things I learn around here!

  16. GeorgeW says:

    “By the way, apparently ice floes were called flakes by English-speaking arctic explorers until about 200 years ago.”

    An example of semantic shift? Flakes now (at least in AmE) are small and . . . well, flaky.

  17. OED (from 1896):
    6. A (loose) sheet of ice; a floe.
    1555 R. Eden tr. Peter Martyr of Angleria Decades of Newe Worlde f. 305, The flakes or pieses of Ise..doo flote aboue the water.
    1685 N. Luttrell Diary in Brief Hist. Relation State Affairs I. 297 Vast flakes of ice of severall miles.
    1796 J. Morse Amer. Universal Geogr. (new ed.) I. 139 To coast..in small vessels, between the great flakes of ice and the shore.
    1820 W. Scoresby Acct. Arctic Regions I. 243 Immense flakes of ice..resembling fields in the extent of their surface.

  18. Stefan Holm says:

    Happy to hear my English at least being upgraded from absent to obsolete :-).

    Consulting a few dictionaries I found a lot of words in Gmc beginning in fl- (inherited) or pl- (borrowed) which all seem to ultimately boil down to PIE *pele (with derivatives *plat and *plak), all with (vague) etymologies of ‘flat’, ‘broad’, ‘spread’, ‘surface’, ‘ripped off’ (from a surface). Among the English ones are field, flake, flat, flaw, floe, fluke, place, plaice, plain, plane, plant, plat, plate.

    The Danish, German and Swedish dictionaries I browsed show a similar diversity of words and meanings. Another can of worms for scholars in comparative/historical linguistics, etymology and semantics to sort out?

  19. David Marjanović says:

    Oh, German has both rennen “run” and rinnen “flow” (said of small quantities; fließen applies to larger ones).

    And that’s just the beginning of the confusion. In Austria, gehen means “walk” and laufen means “run” (rennen being more colloquial and more specifically fast). In Germany (I’m sure I oversimplify, but I don’t know by how much), gehen means “go”, and “walk” is laufen – so that “aisle on a bus” mutates from Gang to Laufgang!

  20. Stefan Holm says:

    rennen being … more specifically fast .

    You mean as in Hahnenkammrennen ‘cockscomb race’ in Kitzbühel, Austria, the world’s most prestigious downhill competition (for which there strangely seems to be no Eng. Wikipedia entry although several Americans and Canadians have won – plus one Brit back in 1931).

  21. Not a top level page, but the bulk of this one.

  22. dainichi says:

    Yeah, I definitely remember being surprised when I learned that “laufen” in German can mean “walk”, when the cognate “løbe” means “run” in my native Danish.

    This made me think of the danish term “rød løber” (lit. red runner), which is the common Danish term for what in English I know as a “red carpet”. But then I google “red runner” and learn that apparently this can also mean a red carpet!

    I just noticed. In Chinese, the morpheme 走, zŏu, means walk/go, while in Japanese, the borrowed 走, sō, is associated with running. Not sure what shifted what way, but it appears this might not be an uncommon semantic shift cross-linguistically.

  23. “walk” is laufen –

    Except when it’s spazieren, right? Or wandern. Well, that’s not walking, it’s hiking. On the other hand, in British English hiking in the hills can be hill-walking. I think. Oh, and dainichi, I believe that the English cognate of laufen and løbe is leap, which means something else. Jumping (springing, like the Springer in chess, only without a horse).

  24. In East Belgian, bicycle racing is called “wielrennen”. I tell them they’re doing it wrong, but they won’t listen.

  25. the English cognate of laufen and løbe is leap

    Also elope.

  26. Stefan Holm says:

    rød løber

    Has that got a just as straightforward pronunciation as rød grød med fløde? (Sorry, dainici, couldn’t resist, although it’s no problem for me to copy).

  27. I just noticed. In Chinese, the morpheme 走, zŏu, means walk/go, while in Japanese, the borrowed 走, sō, is associated with running. Not sure what shifted what way, but it appears this might not be an uncommon semantic shift cross-linguistically.

    For whatever it’s worth, the original meaning of zǒu in Old & Classical Chinese is ‘run’. Japanese preserves this, though it somehow shifted to ‘walk’ in Mandarin (& in Cantonese, I think, though I believe tsáu 走 might still mean ‘run’ in Taiwanese).

    Is there any consistent direction of semantic drift for the Germanic run/walk words?

  28. David Marjanović says:

    This made me think of the danish term “rød løber” (lit. red runner), which is the common Danish term for what in English I know as a “red carpet”. But then I google “red runner” and learn that apparently this can also mean a red carpet!

    While not applied to the red carpet (roter Teppich), Läufer does refer to long, narrow carpets in German.

    Except when it’s spazieren, right?

    That’s going for a walk. :-þ

  29. I definitely remember being surprised when I learned that “laufen” in German can mean “walk”,

    Depending on the context, a use of “laufen” can be ambiguous as between “run” and “walk”, even in German, and even when some of the available non-linguistic context (never is it completely available) is taken into account. That is, in such cases the hearer cannot be quite sure what is meant – a case can be made for both interpretations.

    But that’s no big deal, since a) it may not make much difference what is meant, in the non-linguistic context, and b) if the hearer imagines it might make a difference, he can always ask what is meant.

    I daresay this is true for any language, although that doesn’t make it so.

  30. Jean-Michel says:

    For whatever it’s worth, the original meaning of zǒu in Old & Classical Chinese is ‘run’. Japanese preserves this, though it somehow shifted to ‘walk’ in Mandarin (& in Cantonese, I think, though I believe tsáu 走 might still mean ‘run’ in Taiwanese).

    Not a Cantonese speaker, but digging around suggests that 走 means “run” in Cantonese, while “walk” is 行 (another Classical holdover now largely gone in Mandarin, although it’s found in certain compounds/set phrases).

  31. GeorgeW says:

    I am not clear on what such an encyclopedia would include. Are there words that have NOT experienced semantic shift over time, within a language or across language groups?

  32. GeorgeW, the dictionary would not be of words that have experienced semantic shifts. It would be a dictionary of semantic shifts themselves. You could use it to look up that, for example, words that mean “run” have sometimes shifted to mean “walk”.

  33. David Marjanović says:

    “walk” is 行 (another Classical holdover now largely gone in Mandarin, although it’s found in certain compounds/set phrases)

    The one that comes to my mind immediately is 不行 bù xíng “doesn’t work”, which overlaps widely in meaning with geht nicht in German.

  34. & of course also 步行 bùxíng ‘walk’ (which I mention only because of its identical pronunciation).

  35. marie-lucie says:

    I had a brief look at this online work, and to my mind it falls far short of what could be called an “Encyclopedia” of anything. It is obviously a preliminary work, a list of individual words with two meanings, one of which may be derived from the other (since the direction of change is often difficult to trace). Since the words are listed in English, the list is obviously influenced by the semantic structure of English: it is well-known that the range of meanings of a given word or apparently basic concept is not the same in every language. A good start, but it needs to be much better organized. It seems to me that a “cloud” organization and display (if I understand the concept rightly) according to general concepts, although perhaps more difficult to implement, would be more appropriate and useful.

    Here I submit my interpretation of semantic structure and change as “amoeba-like”: meaning is not (or only rarely) a combination of + and – features, and semantic change is not a matter or adding or substracting such fixed features, but a much looser structure within a flexible and often porous boundary. Thus the meaning of a word can change over time, extending or contracting a boundary, merging partly or wholly with another meaning, splitting into two or more different meanings, changing its exact content, etc. That’s why attempts to describe such losse structures and the changes they are prone to within a strict, rigid formal framework have not been successful.

Speak Your Mind

*