Last week (May 27, 2020), I discussed two attempts to solve the etymology of sword. The second of them would not have deserved so much attention if Elmar Seebold, the editor of the best-known German etymological dictionary, had not cited it as the only one possibly worthy of attention. His is a minority opinion, which does not mean it is wrong, though I believe it is.
Whenever we try to reconstruct a linguistic fact, it is reasonable to follow the model of widening circles: first look for the source nearby, then a little farther, and so on, and, if all such clues fail, suggest a borrowing. Although this procedure does not mean that a borrowing is in principle less likely than a source round the corner, the detective procedure suggested above is logical. (In the Russian fairy tale of the frog princess, three princes were told to shoot arrows and bring wives from the places where the arrows would land. As usual, the youngest brother had bad luck: his arrow fell in a nearby swamp, and he had to marry a frog, but that frog turned out to be a real princess. Hence the moral: explore the home swamp before going abroad.) That is why I ended the previous post with the question about the Germanic (rather than Luvian, Sanskrit, Greek, or Latin) words that resemble sword.
However, before we begin our search, it should be remembered that in Engl. sword, w was at one time pronounced (as also in two, for example: compare twain, twice, and twin) and that the vowel o is secondary; it arose under the influence of that very w. This may be called the Cheshire cat effect (the creature is gone, but its smile remains); that in German, s yielded sch (pronounced as Engl. sh): hence Schwert; while in Dutch, this s was voiced: hence zwaard. Finally, the ancient final consonant was ð (= Engl. th in the). Thus, the Germanic protoform of sword (without Gothic, where the word did not occur) was swerð-; the hyphen stands for the ending that does not interest us here.
English sword has one look-alike, namely, sward, and we’ll return to it later. German has schwer “heavy” and Schwär(e) “ulcer, festering sore, abscess” (related are Geschwür, the same meaning; schwären “to fester,” and the much more familiar schwierig “hard, difficult”). However, schwer, with cognates in and outside Germanic, had a long root vowel in the past and cannot be related to swerð-, because short e and long e (the latter from long æ), do not alternate by ablaut.
In contrast, Schwär(e) turned out to be a promising candidate for being a member of the sword family. The Schwert ~ Schwäre comparison suggests that the sword got its name because it was a weapon that caused pain by incurring wounds. This explanation of the word was adopted in the most influential dictionary of Indo-European, but in 1932, Willy Krogmann, at that time a young scholar, explained that the idea of “pain” must be traced to “piercing,” as, for example, in Engl. bitter from “biting.” We can still find Krogmann’s explanation in the dictionaries that venture to say something about the distant origin of sword beyond listing cognates (the most common case). The objections to this etymology (sword as a piercing weapon) are several. Two are worthy of mention: the function of the suffix –ð (from –ða) remains unexplained, and, if sverð– meant “piercer, cutter,” it is rather strange that the word was neuter, rather than masculine (it is still neuter in Dutch, German, and Scandinavian).
I am now coming to the culmination of my story. Etymologies are seldom impeccable, but one should never ignore small complications. Here is one of them. Dutch zwaard means “sword” and (!) “leeboard”; middenzwaard means “centerboard.” No one doubts that zwaard1 and zwaard2 are two senses of the same word (“sword” and “board”), rather than homonyms. How could that happen? In 1915, Hans Sperber, an active etymologist (among many other things), published an illustration of a wooden club with a sharp side, used in South America as a kitchen board for cutting ham and as a potential weapon. In some ways, it resembles the blade of a sword. Sperber stressed the likeness between two forms—a side of bacon and a slab of timber cut from a tree trunk—and noted that both can be called flitch in English and baco in Swedish (baco is related to Engl. bacon). He believed that the objects once called swords first functioned as boards for cutting meat, and referred to German Schwarte “bacon rind, crust.”
Sperber’s hypothesis has almost never been discussed. Jan de Vries, one the most distinguished philologists of the twentieth century, mentioned it in his etymological dictionaries of Dutch and Old Icelandic and suppled the reference with an ironic exclamation mark (read: “Can you imagine such nonsense?”). Several years ago, our contemporary expressed doubts about the development from “a side of meat” to “a side section of a piece of wood.” However, no one until very recently has returned to a possible connection between Schwarte and Schwert. Only the Russian and Ukrainian researcher Viktor Levitsky discussed it at length, but his conclusions about the origin of the root are beyond the subject of this post. Note: in English, the corresponding pair is sword and sward. I promised to return to sward, and this is the time to do so.
The two words (sword and sward, from swerð– and swarð-) are related by ablaut, like, for instance, German sterben “to die” and its past tense starb. At one time, Engl. sward meant “skin of the head.” Perhaps the word is a borrowing from Scandinavian. In any case, its initial sense must have been “a smooth surface.” Engl. sward and greensward (see the header!) refer to an upper layer of the earth.
The origin of sward ~ Schwarte is allegedly “unknown,” but at least two things about it are known: sward is, almost certainly, related to sword, and Dutch zwaard means both “sword” and “side of wood, etc.” Middle Dutch swarde meant “hairy skin”; the earliest Dutch forms of the word for “sword” were spelled as swart and swert (with final t going back to d). They looked like twins even then.
Where the paths of sword/sward and German Schwert/ Schwarte crossed remains unclear. We may be dealing with one ancient word that split into two or with two close synonyms in different grades of ablaut which influenced one another. Both seem to have designated an object having a smooth surface. It is therefore not unlikely that sword first denoted a utensil having this characteristic, and that later the name was transferred to the weapon. The sword, it appears, was called “sword” because it was smooth and (?) hence shining.
The oldest fanciful etymologies of sword are rather numerous. Only one deserves mention here. It has been suggested that Latin sorbus “service tree” (a kind of mountain ash or rowan tree; Engl. service in its name is an alteration of sorbus) is a cognate of swerð-. This derivation assumed that sorb- (from sworb?) had once designated a spear and a sword. Such cases are rare but known: one example is Old French glaive. Yet the existence of wooden swords finds no support in archeology. Since Sperber derived sword from the name of a wooden board, he shared the sorbus idea. There is no need to defend it. Also, the origin of sorbus is unknown, and one word of undiscovered etymology should never be used to explain the history of another obscure word. The path I suggest was from an object with a smooth (and shining) surface to the weapon. The distant origin of the root should remains a matter of conjecture.
Feature image credit: Photo © David Wright. CC-by-SA 2.0 via Geograph.uk.
Dear Prof. Liberman,
Your explanations on the processes that etymologists use in their research are fascinating!
In this blog and others you mentioned “gender” as one clue in figuring out if look-a likes are really cognates.
How stable is gender between languages? If a word is masculine gender in one cognate, but neuter in the other, could they still be used as a piece of evidence for a relationship?
All the best!
Ablaut example from Latvian – svērt (weigh) – svars (weight) – svira (lever) though has nothing to do with a sword (zobens < zobs (tooth). Still I would suggest to consider usage of sword and the fact that sword fighting requires at least two of them which are kind of in balance until one gets the upperhand. Usage and knowledge of swords might have gone forever.
“The objections to this etymology (sword as a piercing weapon) are several. Two are worthy of mention: the function of the suffix –ð [th] (from –ða) remains unexplained, …”
The most likely explanation is this –ð [th] comes from the Greek “σπαθη” (sword, blade). Attested in Ancient Greek. And giving rise to English “spade”.
I am claiming it is also the distant past origin of “sword”. All these words have the common meaning of a sharp wide cutting object. Whether a weapon or a gardening tool.
Anatoly, you threw at this etymology of “sword” everything but the Greek sink! Thus missing the truth.
You can do better!
A note on Schwert Schwarte:
According to Levitsky (at least in “Germanische Bezeichnungen für „Schwert“ und semantische Typologie” , 2009, the pdf is available online) the tertium comparationis here isn’t “smooth and shiny surface”, quite the contrary: he relates both words to the root *suer “to cut” on account of their coarse surface – on which one might cut one’s fingers I suppose – which suggests that schwarte originally meant something that is raw to the touch, like “rind” or “crust”.
Sperber also wonders whether Schwert might be derived from Schwarte meaning “rind” (rather than “bacon”), but not due to its coarsee texture: he conjectures that the earliest swords were not made of metal but rather wooden – cut out from rind boards (German “Schwartenbrett”).
“…and referred to German Schwarte “bacon rind, crust.” If one tries to explain Dutch ‘zwaard’ (both the weapon and the maritime term), one should also mention ‘zwoerd’ (bacon rind).
Comments are closed.