Oxford University Press's
Academic Insights for the Thinking World

Questions and answers: January 2018 etymology gleanings

The most ancient roots

The question concerned the root – that is said to underlie the English words oar and row. Where did the root come from? This question is almost equal to the more basic one, namely: “How did human language come into being?” The concept of the root is ambiguous. When we deal with living languages, we compare words like work, works, worked, rework, worker, and the rest and call their common part their root. The botanical metaphor in linguistics (root, stem, branch, etc.) is old. But when we open a historical dictionary, we are given the impression that the roots existed before the words that contain them, that roots generate words. – looks fine as the root of the ancient forms of oar and row, but did this egg come before the chickens we can observe?  Only sound-imitative and sound-symbolic roots are “real”: oink-oink, bang, kap ~ kop, and their likes. The question about the origin of – cannot be answered with certainty. Perhaps it indeed existed and imitated the sound made by an oar in water, but we cannot know.

Triumph of the botanical metaphor. Image credit: A simplified language tree covering the Germanic languages, intended for elementary school by Ingwik. Public Domain via Wikimedia Commons.

The origin of the verb to have

This verb has exact cognates in all the Germanic languages, but initially it did not mean “to possess.” For “possession” Germanic used the verb that in fourth-century Gothic sounded as aigan, so that Modern German eigen and Scandinavian eiga (as in Icelandic) are its closest relatives. Their English congener (owe, from āgan) has changed beyond recognition, and its checkered history need not concern us here. The adjective own shares the root with the homonymous verb. Another descendant of Old Engl. āgan is ought; its connection with owe has been forgotten.

The oldest form of the verb have was haban (so, for instance, in Gothic). Two clues to its origin exist. Clue No. 1. Next to haban, Gothic had the verb hafjan “to lift” (its cognates are Modern Engl. heave and German heben). I will state the next fact without any explanation, in order not to clutter the story with cumbersome details. Hafjan is a derivative form, traceable to haban. One can expect some connection in meaning between the initial and the derivative forms, but a tie between “lift” and “possess” is weak. This is the first indication of the fact that Germanic haban did not mean “to possess.” Clue No. 2. In dealing with Germanic, we always try to find a word’s cognate outside the Germanic group: some related word in Sanskrit, Greek, Latin, Celtic, Slavic, or elsewhere. Here we immediately notice Latin habeo “to have.” But our joy is premature. According to the rule of the First Consonant Shift, Germanic h should correspond to non-Germanic k. For instance, the Old English cognate of hwæt is Latin quod (that is, kwod). Both mean “what.”

Oink-oink: an ideal etymology. Image credit: “Pig Alp Rona Furna Sow Happy Pig Animals Nature” by Mutinka. CC0 via Pixabay.

Before the discovery of such basic rules as the rule (or law) of the First Consonant Shift, look-alikes constantly seduced language historians. We now know that too much similarity can be detrimental to establishing affinity. Though a possibility exists that in such cases we are dealing with a borrowing, it would be rash to posit a loan of such a basic word as have. Also, as I have noted more than once, if we suspect that a word wandered from one language to another, we have to explain why a borrowing was needed and under what circumstances the process took place.

To return to our law: for h in haban we need k outside Germanic, and for b (which in this word designates v, from f) we need p. Latin obliges us at once with the verb capio “to catch, seize, grab.” Habere remains an odd man out, and I’ll return to it in a moment. The etymological match between haban and capere suggests that the verb’s initial meaning was approximately “to catch.” The connection between haban “to catch” and hafjan “to lift” poses no difficulties. Apparently, the recorded meaning of haban and its Germanic cognates is the result of a later development.

Latin habeo is not only an odd man out but also a thorn in the flesh. Granted, haban and habeo are a bad phonetic match, but their phonetic and semantic proximity (actually, near-identity) is almost too good to be untrue. That is why many moderately successful attempts have been made to connect them. Similarities between Latin and Germanic are many. The same is true of their legal systems. It is not impossible, even though not very probable, that, while the two tribal unions remained in close contact, the original meaning of haban fell under the influence of its Latin near-homonym. Or there may have existed a common (“Indo-European”) verb beginning with gh– that was conflated with the reflexes (descendants) of the kap– forms. All this is intelligent guessing. Only one thing is clear, the earliest Indo-European kap– verb did not mean “possess.”

Hemingway predicted our dilemma but did not solve it. Image credit: 1st edition cover via The Antiquarian Booksellers’ Association. Fair use via Wikimedia Commons.

Mad as a hatter

I have now read the comment by Mr. Pascal Tréguer concerning the idiom mad as a hatter (the post for 24 January 2018). He too expressed doubts about the connection between this idiom and mercury poisoning, and what he said was partly new to me. I did not know that the reinforcing phrase like a hatter had at one time been used so broadly in northern and Scottish dialects. I am grateful for his remarks, especially because they led me to his blog “Word Histories,” of which I had been unaware. However, let me repeat: I have no chance of discovering the comments added some weeks, months, and years after the appearance of a certain post. This time, I was licking my plate clean for “questions and answers” and went a month back. Those who have late comments are tearfully asked to alert me to their existence in some way or another. The discussion about a mad hatter is for the moment closed. Those who disagree with my reasoning are welcome to believe in the now current theory.  And no, I don’t use this phrase and never heard it from anyone.

The German family name Fartle

A German family having this name arrived quite some time ago in America and for obvious reasons did not want to retain it (they changed it to Hartle). The question was about the name’s origin. I have several very full explanatory dictionaries of German surnames, but none of them mentions Fartle. The parts of my own suggestion, unfortunately, do not cohere. The –le in Fartle made me think that those people had come from Swabia or some other southern part of Germany, and I thought that the root meant “fart” even in their homeland. The German for “fart” is furzen. Its z is the product of the specifically German change known as the Second Consonant Shift (compare Engl. to and German zu), so that t in Fartle poses no problems, but only assuming that the name is northern (the German shift loses its force around Cologne/ Köln). Thus, I ended up with a northern German root (t for z) and a southern suffix. Nor is the vowel too good for my conjecture: a is English, as in fart; Swedish has fjärta and dialectal fräta, and I could not find a German dialectal form of this verb with the vowel a. But that obstacle is not fatal. If some of our readers know the origin of the puzzling name, their response will be welcome.

German ß

I did not quite understand the comment on my remarks. The pair (ver)ließen ~ (ver)lasen probably has different consonants everywhere in Germany (voiceless ~ voiced, strong ~ weak, or whatever). Given the traditional German spelling, ambiguity arises in words like Fluß “river” (the old spelling) versus Fuß “foot,” because Fluß has a short vowel, while in Fuß, u is long. Foreigners confronting words like Ruß “soot” cannot know how to pronounce it. For consolation, I may add that foreigners coming across the English word soot don’t know the vowel length either: oo designates a long vowel in mood, but a short one in foot and good.

Finally, I would like to thank our correspondent from Hong Kong, who assured me that, even though I receive relatively few questions, many people read the blog.

Featured image: Hong Kong, we appreciate friends from far and near. Image credit: “Hong Kong China Night Cityscape Coastline Coast” by Free-Photos. CC0 via Pixabay

Recent Comments

  1. John Cowan

    I don’t quite understand your examples either. TLet me start over.

    In general, one of the ways German has to encode the shortness of a vowel is to double the consonant following it: Rad has a long vowel, Ratte a short one. But this will not work for s, because ss is coopted to mean something different: /z/ is written s and /s/ is written ss (except in Southern varieties where [z] does not exist, much less /z/). The distinction is neutralized in final position and before consonants, so only history tells us to write Finsternis rather than Finsterniss.

    Historically, /s/ had two different spellings, ss and ß, originally a ligature. The distinction between them was typographical: if a syllable break was possible between the two esses, then ss was used, otherwise ß. (This is a special case of the general rule that ligatures are not divided between syllables.)

    The spelling reform of 1990 abandoned this principle, and coopted ß to mean that the preceding vowel was long (or a diphthong), a distinction not representable in the old orthography. Thus whereas before we wrote daß, because in final position there can be no syllable break *das-s, now we write dass because the vowel is short.

  2. Rudy Troike


    The discussion about the German family name made me wonder whether it was real or fictional, like many of the “just so” etymologies floating around on the Web.
    I did in fact have a classmate named Nancy Kneiple, who was careful about insisting that the “K” should be pronounced (and another named Peggy Passwater, who was unapologetic about her name, though I wonder whether it really originated from the inferred meaning). Somewhat different was Dick Raper, who liked to joke that he came from a long line of rapers.
    Could you cite a German dictionary which is devoted to the etymology of family names?

  3. Rudy Troike

    On “have” vs “habeo”, the fact that Western Europe is a well-known Sprachbund, and that Latin texts were long read (and sometimes translated) througout the area. opens the possibility that literacy could have been a channel of influence on the use of this “false cognate”. Speculative etymologies in many dictionaries are full of examples of “perhaps influenced by” a phonetically or semantically similar form.

    More profound, however, is the striking parallelism in the use of Spanish “haber” and English “have” in the formation of the Perfect aspect (alas, I don’t know Latin or French), since this iis a deep matter of a basic grammatical construction. Spanish, however, uses “tener” for the full verb sense corresponding to English “have”.

    (The old alliterative doublet “to have and to hold”, like the repetitive “time and tide”, “hearth and home”, though relatively late, does point to a connection of “possess” with “lift”, as one — folk etymologically — must hold in order to lift.)

  4. Constantinos Ragazas

    Anatoly, you write

    “The question concerned the root rō– that is said to underlie the English words oar and row. Where did the root come from? ”

    Answer: The word “rō” is a Greek word that means “flow”. It is well attested in Ancient Greek texts. How far back do we need to go?

    But if we must, then I suggest the “rolling r”, also in Greek, which suggest rolling movement of water.

    As for “mad hatter”, your logic for your position is deeply flawed, regardless of the etymology. The symptoms of “mercury poisoning” clearly would predate the identification of the desease. Why you don’t see that?

    Of course, by expanding the idioms to “like” and the like, you can always obfuscate the matter in a fog of scholarship.


Comments are closed.