But when you owe, you’re under an obligation. The obligation is most readily thought to be financial or transactional, but one way the word might change is for the sense of obligation to become more general. There was a time when ought was often interpretable as meaning either “owed,” the basic sense, or “obligated,” in a more general one, as in a twelfth-century passage about God (with the Early Middle English rendered as Modern): “For he made all creatures, and nevertheless the world didn’t know him as it ahhte,” with ahhte at this time translatable either as “owed” or “ought.” The world didn’t know God in the way that it owed it to him to do, which is close to saying “as it ought.” After a while, ought came to refer unambiguously to what you should do in a general sense, which alongside words like may, must, and should, is a grammar word. Meanwhile, a new regular past tense form using the -ed ending, owed, took the place once occupied by ought, which had now moved on to, as it were, grammar pastures.
“Can you hear what I hear?”—now sing to the same tune “Can’s a piece of graaammar…” As grammar, it presumably started as something else, and it did: cunnan in Old English meant “know.” Ben Jonson in The Magnetic Lady has Mistress Polish praise a deceased woman for the fact that “She could the Bible in the holy tongue.” We can’t help at first suspecting a typo—she could what? But could meant, all by itself, “knew.” There was even an old expression “to can by heart” alongside our familiar “know by heart.” Modern English is littered with remnants of that stage: other offshoots of cunnan are cunning and canny, all about having your wits about you. Plus, the past tense of cunnan was a word pronounced “coothe,” from which the couth in uncouth comes: the uncouth person is lacking in know-how, as in the kind that lends one social graces.
But if you know, it follows that you are able, and along those lines, can voyaged from the concreteness of knowing to the abstractness of indicating general ability: Yes, We Can!
Or one more: Let’s go! What are the words in that sentence? Technically, let, us, and go. But are we really using the word let? If we are asking to be let to go, then who was holding us back? Let go of me! Let it alone! Let it run down! We can’t say we are using let in those senses when we say, Let’s go, nor are we really saying us when we say the ’s part. Spelling, as always, deceives: the spelling of let’s makes it look like a contraction of let and us, which it once was—but things have changed. While it feels perfectly natural to say is not instead of isn’t, notice that saying Let us go feels like we’re doing bad Shakespeare. In barely any real-life situation, no matter how careful a speaker one is, would one say, “Let us go.”
That’s because let’s is no longer let + us. Let’s is a piece of grammar. It refers not to the action of letting, but to the function of encouragement. Lowth puts it, “Let does not only expreſs* permiſſion; but praying, exhorting, commanding.” Exactly: all languages have a way of rallying the troops in that way. It’s a command applied to us rather than someone else, a musty but useful term for it being the hortative. Languages like Latin and the Romance languages do it by putting a verb in the subjunctive: ¡Viva Zapata!, “May Zapata live!” in which the -a ending on vivir signifies the subjunctive mood. Other languages have a little separate word to do it, or a suffix, or any number of strategies. English’s is a little grammar word that by all rights should be spelled “letz.”
When Words Lose Their Selves: Grammaticalization Goes Further
Notice that I am sneaking that term in. Grammaticalization is what linguists call it when a word becomes a piece of grammar. I have avoided using the word until now because it’s tempting to think it has something to do with grammaticality in the sense of proper phrasing. It also sounds, with its seven clattering syllables, generally intimidating, or at least like some kind of problem. Some use grammaticization, but that isn’t much better. Yet, here we are, and there’s certainly no better term I can think of myself.
The grammaticalization process can go further than ought and let’s. When a word no longer has meaning per se, and is more something used to organize the words that do have meaning, it no longer gets accented much. You used to be the only one: the melody is “YOU used to be the OWN-ly ONE.” The used to, and the other grammar word the, are mumbled compared to the you and the only and the one, the “word” words. (Be is a whole story in itself, not as much a “real” word as we have reason to suppose.) You could possibly say, “You USED TO be the OWN-ly ONE,” but that’s only for emphasis, the unusual case.
Now add the fact that grammar words tend to be used an awful lot. You can’t talk without using the grammar to put words together, and so used to, the, and a get much heavier wear-and-tear than umbrella, squirt, or avoid. So: we have a word usually mumbled rather than accented, and used that way day in and day out by countless people over vast periods of time. Two things can happen to a word like that.
First, it starts sounding different, losing the ringing, robust sound of solid words like goat, jolly, and deodorant. Second, in that weakened state it starts leaning, as it were, against a robust “real” word next to it. After a while, it gets accustomed to that and can no longer stand up on its own at all. Or, to step back from the anthropomorphizing, when something is said without an accent and uttered constantly, we can start to hear it as a syllable of a word next to it, rather than being its own word. Think of a toolshed. We don’t think of it as a building, precisely, even though it technically is. Small, uninsulated, of narrow purpose, and ugly, it’s an appendage to a house, an architectural punctuation, even if it isn’t actually adjacent to the house. You’d never build a toolshed by itself out in the middle of a field, and in fact, one might be nagged by a sense that if a toolshed isn’t right up against the house, it kind of should be. It’s not really a building, after all. In fact, a garage might seem a better bet—keeps everything nice and connected.
A heavily used little grammar word has a way of becoming a toolshed. After a while, it actually is a syllable of other words, and can never stand alone—the birth of a prefix or suffix. A good example is the -ly that forms adverbs like slowly and gently. It started as the word like (pronounced “leek” back then). That’s easy to imagine because even today we can still say slow-like to mean “in a slow fashion.” Used that way constantly, however, like lost the accent it once had. You hit the like in slow-like fairly hard, but the -ly in slowly not so much. Mumbled, it lost its final consonant—“leek” became “lee.” And in that state, where it no longer even sounded like what it had originally been, it became easy to hear this “lee” as something hanging off slow rather than its own word, as something hanging off warm, quick, light, etc.—until now there became an adverbial suffix -ly that no one would ever associate with like except scholars of language change who know where the bodies are buried. (As it happens, like originated as a word meaning body!)
What becomes a word, then, can wind up as a mere syllable. The suffix -hood in childhood, motherhood, and neighborhood sounds familiar in its way, because there is a word hood. We might think that these words are simply doubledecker words like blackboard and cheeseburger. However, the word hood as we know it doesn’t mean “condition” as it does in these longer words. If what we know as hood were the source, then neighborhood would refer to the person next door’s shawl, an unlikely thing to have a word for at all. And as for the now well-known slang term ’hood for neighborhood, clearly that term arose long after childhood, motherhood, and the others. No one was calling themselves “from the ’hood” during the Iron Age.
The -hood suffix is actually from an Old English word pronounced “hod.” It meant “condition” or, by extension, “manner” or “way.” In Beowulf, Grendel gets hurt through hæstne had—“violent means,” as in “in a violent way.” However, used next to words like child, mother, and neighbor, hod morphed into -hood. Today’s term the ’hood technically means, then, “the condition”—go figure.
This kind of story is in the past of most suffixes and prefixes. The -ed that marks th
e past probably started as did: painted goes back, in other words, to paint-did. Considering how often you refer to things in the past, you could have known that did, when used that way, was eventually going to wear down to something like an -ed, inseparable from the verb itself.*
Didn’t You Use to Be Somebody? When the Journey Ends
The life cycle I am depicting can include, as you may have suspected, that stage we don’t like to talk about: demise. When what is now just an appendix to other words, unaccented and barely thought of as language, is uttered incessantly enough over a long enough period of time, it can just plain wear out and fade away.
This is what happened with English verbs, to such an extent that the only signs left of what once was are peeking out from behind the curtains of the language, as unnoticeable as the dust specks in the air that you catch sight of only when the light is slanting in just the right way.
Here are some Old English verbs. You don’t have to know or care a thing about Old English to notice that all the verbs have one thing in common:
shake
ahrysian
rattle
hrutan
roll
wealcan
wash
wæscan
rinse
aþierran
repeat
geedlæcan
It’s that they all end in -an. All verbs in Old English did: the -an (or sometimes -on) was what marked something as a verb, just as -ly marks something as an adverb today, but much more consistently. (The ending -ly is used often, but not always, as we know from adverbs like often and always!) Readers who have studied German will be familiar with the way verbs in that language end in -en; Old English, sister language to German, had the same feature.
But the -an was unaccented: the word for rattle, hrutan, was pronounced HROO-tahn. Or, after a while, more like HROO-tun, with the vowel sound in the -an ending drifting from ah into the grubbier kind of sound such as the o in lemon. As time went by, the -n dropped off and left just that little vowel alone. It was even more vulnerable to the elements than when it was protected by an -n, and during the Middle English period it wore away completely. From then on, English had no ending to show that something was a verb.
Today, the old -an ending hangs around on the margins, but it’s something that has to be pointed out, like showing someone that deer standing just over the hill fixing to bolt away. Because we have the words length and strength, we can glean that the -en in lengthen and strengthen serves to make words into verbs. There is also a reason that in older poetry you can catch ope for open. Ope was the natural result of the ending wearing away, but by chance, the old open stuck around, kind of like that guy in college who kept visiting after he graduated, maybe a little too often, and after a while got a job on campus and, last you heard, is still there. Open was like him: you just couldn’t get rid of it, and it ended up becoming the standard form. However, we don’t think of the -en in open as a suffix, any more than we think of the -en in listen as one. Listen is another example of the old form happening to hang on—by all rights it “should” be just list, and you can actually find that word in, again, old poems.
But the ending remains a dead thing. We can’t make new words with it. Today, even if you are one of the people who reviles the use of nouns like structure as verbs, you certainly aren’t wishing people would say “structuren” instead, and I don’t recall anyone giving “faxen” a try instead of just saying they needed to fax something. When the founder of The Simpsons’ Springfield coined the town motto “A noble spirit embiggens the smallest man,” the use of -en where it isn’t already established can only register as funny.
Grammaticalization Is Something Happening, Not Something That Happened
A natural question at this point might be: if endings are always wearing away like this, then how does a language keep any grammar around at all? There are two answers.
First, the endings (and prefixes) are not always wearing away. The prefixes and suffixes that are less necessary to making sense are more vulnerable than others, but when there is a bit of stuff that we really need in order to express thoughts, it tends to hold on tight. The Old English verbal ending -an was low on the scale of necessity; it’s usually pretty clear that something is a verb from the fact that it refers to an action. However, English is in no hurry to let the past-marking ending -ed wear off, because one achingly needs to refer to the past and, in fact, does so more than one refers to the present or future.
I feel bad for the teacher my elementary school hired to give us Spanish lessons in fourth grade. She deserved major props for actually teaching us the present tense conjugation instead of just a scattering of nouns and expressions, but I kept asking her when we were going to learn to put things in the past, because it seemed to me that every second thing I wanted to say was in the past. But the past in Spanish is harder to teach—I must have gotten on her nerves. Yet within my obnoxiousness I was onto something: you need that past tense to get beyond the “My uncle is a lawyer but my aunt has a spoon” level of things. For example, while it is inherent to casual speech to let some things slide, people make unconscious calculations as to what is expendable. People actually are much more likely to say, “We saw Wes’ Side Story,” letting go of a t that has no meaning, than to say, “Yesterday I kiss’ my daughter” and leave off the -ed that is so central to basic meaning.
Therefore it isn’t that every grammatical bit that isn’t accented is ripe for elimination. If you are Jewish and went to Hebrew school, remember being taught (or not!) that you change the verbs’ tense by changing their vowels, so that kotev is “writes” but katav is “wrote”? That’s grammar, and it’s been hanging on for thousands of years. It’s a trait of the languages in Hebrew’s family, Semitic, and none of them has just sloughed this stuff off, accented or not.
Beware, in fact, the myth tripping up not only the general public but many linguists, that languages regularly let all their prefixes and suffixes fall off. This conception is based essentially on the fact that Modern English has so much less noun case marking and verb conjugation than Old English, and that Latin lost its noun case marking when it became the Romance languages. Indeed, while Latin nouns were marked for case in this way (stella means star):
nominative
stella
genitive
stellae
dative
stella
accusative
stellam
ablative
stella
you would never know any of this from étoile, the descendant word in French for “star.” The only way you can change it is to plonk on a plural ending -s for étoiles (and much of the time, even that -s isn’t pronounced). In French, case is marked in similar fashion as in English, with prepositions like of and to, and the accusative is all but ignored.
However, this is by no means what happens to languages in general. Around the world, countless languages with case marking like Latin’s basically keep it. The typical language is full of “messes” like this and much more, and has been since time immemorial. Languages like Old English and Latin took it all off because they went through periods where they were learned as much by adults as by children, which makes a language less complex than it would normally be. Old English was never the same after the Viking invaders picked it up starting in the eighth century. Latin changed when imposed upon subjects of the spreading Roman Empire. But these were abnormal circumstances.
Beware, then, the idea that languages regularly go around dropping away the hard stuff and becoming more like English. If this were the case, then wouldn’t all languages actually be like English by now, and since they aren’t, what’s taking them so long? A suffix may wear off, but just as often being a suffix—or a prefix, or a vowel change of the Hebrew kind, or the difference in English between run and ran—is home base, the homestead, The Way It Always Is.
Now, eventually, as meanings change and also sounds change (as we will see in chapt
er 4), even a grammatical bit that serves a needed function starts getting jostled out of existence. A piece of grammar doesn’t last for, say, tens of thousands of years. However, by then the subject of the second answer to our question comes into play.
Namely, new things are always grammaticalizing as older ones wear away. Grammaticalization is not something that happened long ago; it’s something that is happening now, too—part of what it is to be a living language. The same inevitable processes of creeping reinterpretation that make words go personal or just go some other way (as in chapters 1 and 2) are also always at work turning some words into grammar.
The impulse has its roots in the more general fact that, as much as we are condemned for being sloppy and vague in our language usage, humans are deeply committed to explicitness and clarity. In the back of our minds, we are always sprucing the language up, making sure it does the job, like a director regularly working with the cast of a long-running play to keep it fresh. This is the source of words condemned as redundant, such as the famous irregardless. One does not technically need the ir- when the -less of regardless already conveys the negation. However, the fact that the reinforced version with ir- has been so readily taken up indicates that people want that negation component to be as clear as possible. In the word overwhelm, you might wonder why there is no word whelm. In fact, there once was. You might wonder what it meant, and it meant … “to overwhelm.” Overwhelm began as redundant as irregardless; people were simply being forceful.
Think of how often people today say things like sink down and rise up, when the verbs themselves, sink and rise, contain what down and up contribute. The impulse to add down and up is one of clarity: rise up is stronger, more vivid, than rise. One could just do with rise if one were, for some reason, limited to expressing only exactly what was necessary with as little expenditure of energy or verbiage as possible. Yet who set this condition for human language? Note how incompatible such puritanical parsimony seems to almost any other endeavor we value or esteem—what art, what culture, what feeling, what ingenuity, what humanity could spring from such a prescription? Sink down isn’t messy; it breathes, commits, lives. Hence similar constructions one hears, such as penetrate into and separate out: redundant, yes. But it is integral to being a person to be, to a degree, redundant, something in other contexts called energetic or spirited, and what we might usefully term underlining.
Words on the Move Page 10