So the grammarian tells you that you shouldn’t put prepositions at the end of sentences. And you say sure, this guy’s a grammarian, he knows what he’s talking about, I’ll try to do what he says.
But suppose you’re more of the skeptical type. You find something fishy about the claim. You demand evidence. How would you go about fact-checking this? The best evidence comes from observation of native speakers. And if you go do some observation, it turns out that people are ending their sentences with prepositions all over the place. The supposed rule is quite at odds with how people actually talk. Something’s up. Do people have bad grammar? Or is the rule wrong? This is a rhetorical question of course: obviously the rule is wrong. Let’s think about it.
Go back in time. Suppose you had never talked to the grammarian in the first place. Suppose instead that you first spent time listening to and recording English speakers. On the basis of your recordings alone, without the influence of the grammarian’s opinion, what conclusion would you draw about English? We would have to conclude that sentence-final prepositions are acceptable because we observe so many of them, just as we would be led to conclude that nouns follow articles because that’s what the recordings show.
The only way to understand English grammar is to look at what speakers of English do and don’t do. Either you draw your conclusions about English from observations of English, or you’re just making stuff up. And as a side note, I don’t mind when people make stuff up in order to set standards. For example, I’ve nothing against a rule that says sentence-final prepositions should be avoided in formal writing. Problems arise when people start to confuse those standards with actual descriptions of English grammar. Standards are fine, so long as we don’t forget that they are essentially random choices about grammar, generally based on the personal preferences of people who speak the majority dialect.
It might help to think about this with a language you don’t speak. Suppose you were given the time, money and incentive to go record the grammar of a rare (imaginary) language Zoza. The Zoza live in a remote tropical region, do not have schools, and there is no written history of the language. That means there are no grammarians or grammar books to consult. Luckily, a linguist visited a few years ago and recorded a list of a few hundred words in Zoza to get you started.
How could you go about figuring out the grammar of Zoza? There’s only one way: you have to go listen to Zoza speakers and do what they do. What other choice is there? You can only determine how the language works through observation.
Suppose in advance you decide “In Zoza, objects come after verbs”. Is that approach going to work? Only if chance is on your side and Zoza has that kind of word order. But suppose after a few days with the Zoza it becomes clear that actually objects come before verbs. You would have to change your mind about how the grammar works. Not to change your mind would be needlessly stubborn and the Zoza would probably wonder what is wrong with you if you can’t get this right despite daily contact with them. So why is anyone still pretending like English doesn’t have sentence-final prepositions? What is wrong with you? We’re drowning in the evidence for them.
Along this same line, it’s interesting to think about this from the perspective of an infant. A baby doesn’t know which or how many languages will be spoken in the community she’s born into. She’s just learning what her home language(s) look(s) like, one utterance at a time. The only way for that baby to figure out the language of her parents and caregivers is by observation.
It’s even harder than the Zoza scenario, actually. At least in that case you had some words, and you already speak a language. Babies are starting from scratch. They don’t have another language to fall back on. All they’ve got to go on is the behaviour of other more experienced users of the language. And if those people are putting prepositions last, then that’s what babies will infer is correct. Babies don’t have access to some magic oracle that will tell them in advance that people might put prepositions in one place but really don’t do that it’s all a trick.
What really renders this a hard task is the fact that words don’t come with grammatical labels. Babies have to figure out what the grammatical categories are at the same time they learn the language. Nouns don’t come with a silent [noun] sign attached to them. Babies/toddlers/children must learn how to classify words based on how they hear those words being employed. Which is how we should all think about part of speech anyway. By the time a kid is old enough to actually understand the concept of “part of speech” and learns about it in school, it’s WAY to late for it to have any real effect on their language.
When a child infers that sentence-final prepositions are how the language works, then when this baby grows up she will produce such sentences. And her children, should she have any, will be learning this language from scratch too, and what they will hear is sentence-final prepositions, so they will also infer that the language works that way. And before long *everyone* is doing that with their prepositions and there is no longer any evidence available to learners to the contrary.
This is how we must study language too. You can’t assume things about it in advance. We have to learn about it through observation. Anyone who wants to adopt evidence-based approach to their teaching needs to take a hard look at the traditional rules.