4 min
Grammaticalization: When Factual Words Disappear into Function
If you have ever studied a modern language like Chinese or an ancient language like Latin, you may have been surprised to discover that not every language has an indefinite article like a or a definite article like the. English speakers use little helper words like a and the all the time without thinking much about it, but they are strangely abstract—especially when you consider the formation of a language. You can point at a cat or a dog to help make the meaning of such words clear, but there’s no easy way to point to a the. Grammatical words like the are little conceptual luxuries that many languages get along fine without, and in fact English hasn’t always had the definite article the either. So where do such words come from, and how do they catch on? The answer is found in a common process of language evolution that linguists call grammaticalization. This Language Matters article describes the process, offers examples of its contributions to English grammar, and indicates how grammaticalization is still at work today.
The Story of The
As natural human languages develop, lexical words with concrete meanings are often put to work in more abstract, “grammatical” ways. In technical terms, the original sense and sound of such a word will usually get “bleached” or “eroded” away in the process: that is, grammaticalized words tend to lose their concrete meanings and their full phonetics as they take on more conceptual and grammatical jobs. This is how English got the word the. At one early stage in Old English, the demonstrative determiner and pronoun se was used a bit like our modern word that in pointing out particular concrete things. As time went on, though, se underwent sound change to give rise to the Old English þē and eventually the Middle English þe, both of which could just mean something like “the” in a more general sense and thereby ended up giving English the definite article the. The original meaning of se got lost along the way, as the grammaticalized form the slowly took over and allowed English speakers to communicate a more abstract kind of meaning than the original demonstrative “that” meaning.
As it turns out, the story of our word the is not unusual. Other languages have developed definite articles from demonstrative determiners, and similar processes of grammaticalization have contributed to English in many other ways. Here are just a few examples:
Verbal Constructions
When you say something like I will regret this tomorrow, the word will has nothing to do with desire. You may even see the predicted future as happening against your will. The Old English verb willan meant “to wish” or “to want”, and it is the source of the modern word will that carries similar connotations. You can will yourself into motion, or make your tired body submit to your will.
Thanks to the process of grammaticalization, though, the Old English verb willan also got applied to talking about the future. The idea of what people wanted to do and therefore planned to do got folded into predictions of what they were going to do: I will to see you (a statement of desire) thus became I will see you (a statement about the future), and its connection with will in the sense of “volition” was quietly forgotten. Along with the loss of meaning came a loss of phonetics. The future tense will can be used in a contracted form: I’ll get my work done. Compare that with the meaningful will: I will myself to get my work done can never be said as I’ll myself to get my work done. The following examples also exhibit phonetic reduction, if you look closely!
Helper Words
At its roots in Old English, the adverb not is connected to a number of words. A whit was a “thing,” and ā meant “ever,” and so the compound word āwiht meant “anything.” By adding ne (“not”) to āwiht (“anything”), Old English speakers built up the word nāwiht, which meant “nothing” (literally, “not anything”). The latter compound word was shortened to nāht and passed on to Middle English in forms like naught or nought, which could still mean “nothing” but sometimes just meant “not at all.” Modern English sometimes preserves the memory of naught/nought meaning “nothing” here and there (as in the name Dreadnought that refers to a kind of bravery that “fears nothing”). For the most part, though, the legacy that naught/nought carried down from the Old English nāwiht is limited to the grammatical and conceptual word not—a general term of negation.
Affixes
Grammaticalization is the source of many prefixes and suffixes. In modern English, for example, the adverbial suffix -ly in words like slowly and quickly means something relatively abstract like “in the manner of,” but its roots lie in the concrete Old English word lic (“body”). In the evocative sense of the original usage, then, to do something quickly would have meant to embody the physicality of a quick physical thing (i.e. a “living” body or a “speedy” body). In its grammaticalized modern form, the affix -ly is more conceptual than corporal.
A Mainstay That’s Here to Stay
Now that you’re familiar with the phenomenon called grammaticalization, you may still find yourself wondering now and then where some strange English grammatical form comes from, but you can usually be sure there were sensible, and even concrete, reasons for its appearance at one point in the evolution of the language. When frustrating mysteries arise, it’s only because the semantic and phonetic “erosion” involved in grammaticalization has erased the evidence over the ages. The forms that result from this process survive not because they make obvious sense in theory, but because they work in practice. Grammaticalized forms of formerly concrete words take root because they do their job of communicating something. It’s a natural and common process, and in fact it’s still happening.
Consider the word gonna. At one theoretical level, gonna is just a contraction of the two words “going to,” but in practice it’s more than that. Gonna has become a grammatical marker of intention and futurity in its own right. For example, you might say “I’m gonna buy shoes today” if you mean “I’m going to buy shoes today.” You would never say “I’m gonna the shoe store,” though, if you mean “I’m going to the shoe store.” In theory, gonna should work in both cases if it really just meant “going to.” Instead, gonna is taking root as a purely grammatical signal for talking about the future, drifting away from its theoretical status as a simple contraction.
Grammaticalization of this kind is a relatively rare phenomenon, because grammatical words like prepositions and determiners are in a mostly stable and closed class, unlike content words (nouns, verbs, adjectives and adverbs), which get added to the language all the time. Gonna, however, shows that grammaticalization is alive and well, as does the relatively recently grammaticalized conjunction slash as in “a phone call slash informal therapy session.” Apparently, the creative work of grammaticalization will continue (slash is gonna continue?) as long as natural human language does.
