English verbs (and verbs in some related languages) can be divided into two types. One type makes its past tense and past participle by adding a dental suffix, which is pronounced “d” (“I love, I loved, I have loved”) or “ed” (“I want, I wanted, I have wanted) or “t” (“I walk, I walked, I have walked”). The other type makes its past tense and past participle by changing the vowel in the root, as in “I sing, I sang, I have sung”. Verbs in the first group are called “weak verbs” and those in the second are called “strong verbs”—why “strong” and “weak” I don’t know. Most verbs are weak verbs, and new verbs added to English are added with weak forms—“I blog, I blogged, I have blogged”—but the strong verbs are old and persistent.
Many grammar books divide the strong verbs into seven classes. These classes were fairly regular in the older forms of the Germanic languages, but they have become less regular over time. Here are the classes:
Class 1: ride, rode, ridden
Class 2: freeze, froze, frozen
Class 3: begin, began, begun
Class 4: break, broke, broken
Class 5: give, gave, given
Class 6: take, took, taken
Class 7: fall, fell, fallen
The form of a verb stem won’t tell you if it’s strong or weak, and if it’s weak there’s no way to tell which class it belongs to. You just have to learn which is what. The verb “to live”, for instance, looks as if it might belong in class five of the strong verbs, along with “give”, but it’s a regular weak verb (“I live, I lived, I have lived”). It’s no surprise that children just learning English get things mixed up and use the wrong form (“I holded the cup”). People also play around with the forms—So “think” becomes “think, thank, thunk”, by analogy with “drink, drank, drunk”. Dialects can differ as well. Where I grew up the form “brung” was common—as in “I brung my lunch yesterday”— as was the form “swol”—as in “my finger got all swol up”. In North America the past participle of the verb “to get” can be either “got” or “gotten”, but in England only “got” is used and “gotten” is considered archaic.
Modern German also has strong and weak verbs. German weak verbs in the past tense add the dental suffix “-te” and in the perfect add the dental “-t”. Thus the English weak verb “to thank” (“I thank, I thanked, I have thanked”) is matched by the German “danken” (‘ich danke, ich dankte, ich habe gedankt); and the English weak verb “to brew” (“I brew, I brewed, I have brewed”) is matched by the German “brauen” (“ich braue, ich braute, ich habe gebraut”); the English strong verb “sing” (“I sing, I sang, I have sung”) is matched by the Germen “singen” (“ich singe, ich sang, ich habe gesungen”); and the English strong verb “to break” (“I break, I broke, I have broken”) is matched by the German “brechen” (ich breche, ich brach, ich habe gebrochen”.
We can see a similar set of verb forms in ancient Greek. Here are some of the forms of the verb “leipo” (λείπω), which means “to leave behind”, in the present, perfect, and aorist tenses (all first-person singular indicative active): leipo, leloipa, elipon (in Greek letters, λείπω, λελοίπα, ἔλιπον). Greek verbs can be complicated; the “le-” at the beginning of “leloipa” is called a reduplication, and the “e” at the beginning of “elipon” is called an augment, and they have nothing to do with the present topic. The point here is that the roots of these three forms are “leip-”, “loip-”, and “lip-”. That’s reminiscent of the English strong verbs, such as “sing, sang, sung”.
All of these verbs make their grammatical forms by changing the vowel in the root. In a way you can think of the root of “sing” as “s–ng”, with a hole in the middle, which gets filled by different vowels depending on the grammatical form: “sIng, sAng, sUng”. Likewise, the root of the Greek verb “leipo” is “l–ip”, and the hole gets filled in by the appropriate vowel: “lEipo, lelOipa, el–ipon”, with an empty space for a vowel in “elipon”. This process is called “ablaut”—which is the name given to it by the German historical linguist Jacob Grimm (of the Grimm Brothers’ fairy tales). This process goes back thousands of years to Proto-Indo-European, and it shows up, more or less clearly and frequently, in the various Indo-European daughter languages, including English, German, and Greek.
The story of Indo-European ablaut gets complicated, and I don’t want to get lost in the details. For the present post, it’s enough to say that there are three vowels involved—or two vowels and the absence of a vowel. The “e” vowel produces the “e-grade” forms, the “o” vowel produces the “o-grade” forms, and the absence of a vowel produces what are called the “zero-grade” forms. (There are also lengthened versions of the e-grade and the o-grade, but I won’t need them for this discussion.) These three forms are clearly seen in the three forms of the Greek verb “leipo”: the e-grade is “leipo”, the o-grade is “leloipa”, and the zero-grade is “elipon”: “lEipo, lelOipa, el–ipon”. The same system produces the English strong verbs, but with complications along the way: for the verb “sing”, the e-grade is “sing”, the o-grade is “sang”, and the zero-grade is “sung”. Exactly why the e-grade is “i” and the o-grade is “a” and the zero-grade is “u” can be explained, but it’s not important for now.
These examples (“sing, sang, sung” and “leipo, leloipa, elipon”) show various forms within a verb paradigm, but ablaut also shows up in the relation between verbs and nouns. The Greek verb “lego”, for example, which means (among other things) “to speak”, is the e-grade of the root “l–g” and the related noun “logos”, which means “word”, is the o-grade of the same root. There are lots of similar instances of a verb in the e-grade and an associated noun in the o-grade. The e-grade verb “tremo” (“to tremble”) and the noun “tromos” (“tremor”), the verb “pleko” (“to braid”) and the noun “plokos” (“a lock of hair”), the verb “trepho” (“to nourish”) and the noun “trophos” (“a feeder, a nurse”), and so on. In Latin, the e-grade verb “tego” (“to cover”) goes with the o-grade noun “toga” (“toga”).
All the examples so far show vowel gradation within a particular language, but we can also see the different grades of a root in different daughter languages of Proto-Indo-European. Take, for example, the word for the joint in the middle of your leg. The Latin word for this joint is “genu”. The only English word I know offhand with this Latin root—and the only one I can find in my desk dictionary—is “genuflect”, which means “to bend down to the ground with one leg as a sign of respect or worship”. (As a pronunciation footnote, in English we ordinarily pronounce the first letter in “genuflect” as “dj”—think of the “g” in “again” and say the word “genuflect” with that “g” sound and you will clearly hear the difference. In Latin the initial “g” in “genu” is like the “g” in “again”.)
The ancient Greek word for this joint is “gonu”, in Greek letters spelled “γόνυ”. I don’t know of any English word that begins with this root; “gonorrhea” doesn’t count, because it comes from a different root, the ancient Greek “gonos”, which means “semen”, added to “rhein”, which means “to flow” (as in “diarrhoea”). But a related Greek word “gônia” (“angle”, spelled with a long “o” in Greek) does appear in of lots of English words, such a “polygon”, which means “a geometrical figure with many angles”, or “trigonometry”, which is the branch of mathematics originally dealing with three-angled figures, that is triangles. The root “gôn-” refers to angles because if you bend the joint in the middle of your leg you get an angle. (The word “angle”, by the way, is related to the word “ankle”, also because of the bend in the joint.)
It’s not hard to see that Greek “gonu” and Latin “genu” are almost the same, and it’s not a stretch to imagine that they are related: “genu” is the “e-grade” and “gonu” is the “o-grade” of the same Proto-Indo-European root. But we’re not done. There’s another Greek word to add to the story, the adverb “gnuks” (γνύξ: on bended knee), which has lost the vowel to produce the zero-grade form, “gn”.
And that brings us to the English word for the joint in the middle of your leg, “knee”, which uses the zero-grade of the root. The word “knee” is spelled with an initial “k”, which corresponds to the initial “g” in “genu” and “gonu”, but if we just go by pronunciation, the “k” in English “knee” isn’t there at all—the English word “knee” is pronounced “nee”. The “k” in “knee” is left over from an earlier pronunciation; in Old English the word was spelled “cneo”, and both the “c” and the “n” were pronounced. In Modern German the word is “Knie”, and both the “k” and the “n” are pronounced. The Old English “cneo” (and the German “Knie”) lost the vowel, leaving the zero-grade “kn” behind. Then English decided to outlaw initial “kn”, so the “k” was dropped leaving just “nee”—in pronunciation but not in spelling; German, however, allows initial “kn”.
Ablaut explains a lot of English etymologies—for instance, in the topics of my next post, “phlegm” and “sperm”. Stay tuned.
(I would like to thank Friend of the Blog Dr. Robert Fisher for interesting conversations on this and other topics.)