Back when I was writing Fluent Forever, I wrote a chapter about how grammar works and what it does. Ultimately, my editor and I decided to cut it, because it interrupted the flow of the book, and because people tend to dislike grammar, since it can be confusing and finicky.
I think we made the right decision with the book, but still, I liked the idea behind that chapter. I've talked to so many people who feel confused and annoyed about grammar, and I've always wanted to sit them down and explain why grammar is something that's both graspable and beautiful. So I'm going to share that chapter here. I hope you enjoy it!
Note: I'm including a table of contents at the start. The idea of that is so you can come back to this article later on and jump around to useful sections if you want a refresher on what "Cases" are, or what's a "Tense," etc.
Grammar: The Stage Director of Language
A language’s grammar is a kind of stage director, assigning roles to the various words in a sentence, and telling them how to interact.
Let’s take a moment to look at a grammar-less sentence:
SLEEP EAT WORK EAT WORK EAT SLEEP .
This sentence works just fine as it is; you don't always need a stage director to tell simple stories. It’s only when I wish to explain what was eaten or who ate, that I’m going to run into problems without the aid of grammar. If I say ‘EAT COW’ (or ‘COW EAT’), you might be able to infer that I was the eater and the cow was the meal, if only because I was the one who said it and cows are herbivores, but it’s already quite vague. Perhaps I was telling you to eat a cow, or perhaps I was describing a cow eating. As soon as I try something not involving herbivores or myself, like LION ALLIGATOR EAT, we’re both screwed. Who’s doing what? At its simplest level, grammar is all about finding ways to answer that simple question.
I want to walk you through several different ways that languages can answer the Who's doing What question. My goal is for you to learn to step back from the details and get a bird's eye view of grammar's job as a stage director. Once you've done that, then you can step back in and learn your particular language's grammar system with a better sense of perspective. That should make grammar a little less intimidating. Let's start with this sentence:
The man eats the cake.
English is a syntax (word-order) heavy language. We can say sentences like this, but we not can say sentences like this. Going back to our man-cake sentence, we see two players: the man and the cake. Our man is the active player (the subject), and our cake is the acted-upon player (the object).
There are two things to note here:
First, there’s a traditional order to an English sentence, where the verb comes after the subject, and the object comes after the verb. Man [s] eats [v] cake [o]. This is referred to as SVO (subject - verb - object) order, which is the second most common word order, encompassing 42% of the world’s languages (including the Romance languages, Mandarin, Vietnamese and Arabic). The majority of languages, 45%, are SOV (subject - object - verb: The man the cake eats), which you’ll find in Japanese, Hindi and Turkish, for example. Once upon a time, English, too, had SOV structure until the Normans invaded in 1066. We can still see this in archaic phrases like “With this ring, I thee wed” and “Till death do us part”. The other word order possibilities (VSO, VOS, OVS, OSV) are rare, with the rarest being Yoda’s Object-Subject-Verb (“The cake the man eats.”), which shows up in a few languages of the Amazon basin.
Second, English word order is very inflexible. If you mess with the order, you either produce gibberish or a different meaning: “The cake eats the man” paints a very different picture, and “The the eat cake man” means nothing. English is known as a fixed word order language. Other languages ("free word order languages") are much more flexible. Here’s a similar sentence in Russian:
Мужчина ест колбасу.
Mushina yest kolbasu.
(A) man eats (a) sausage.
I can randomize the order of this sentence without causing any problems or changes in meaning (and you can see how much the meaning of the English sentence would change):
Колбасу ест мужчина.
Kolbasu yest mushina.
Sausage eats man.
Eст колбасу мужчина.
Yest mushina kolbasu.
Eats sausage man.
Eст мужчина колбасу.
Yest mushina kolbasu.
Eats man sausage.
Колбасу мужчина ест.
Kolbasu mushina yest.
Sausage man eats.
Мужчина колбасу ест.
Mushina kolbasu yest.
Man sausage eats.
In Russian, you can screw around with word order without losing meaning, because Russian tags each word individually with its role by modifying the endings of each word. A more accurate translation of this sentence would be:
Мужчина колбасу ест.
Mushina kolbasu yest.
Man-SUBJECT sausage-OBJECT eats-VERB-WITH-SINGULAR-3rd-PERSON-SUBJECT
If I wanted to change the story and have the sausage eat the man, I’d just mess with the endings of my sausage and my man, changing Mushina (subject) to Mushinu (object) and Kolbasu (object) to Kolbasa (subject), like this:
Мужчину колбаса ест.
Mushinu kolbasa yest.
Man-OBJECT sausage-SUBJECT eats-VERB-SUBJECT-IS-SINGULAR-3RD-PERSON.
This is a strategy that many languages employ; if you mark words with their roles, then you’re free to do almost anything with a sentence, which gives you expressive possibilities that don’t exist in a fixed word order language like English. If I was telling a this man-sausage story in English, and I had indicated that this man was a vegetarian and I wanted to emphasize how weird it was that he was eating a sausage, then the best I can do is italicize the word 'sausage.' But if I'm telling this story in Russian, I can just bring Sausage-OBJECT to the front. Since Russian is a SVO language, like English, putting the object up front calls attention to it, and so the writer of ‘Колбасу ест мужчина’ has successfully transmitted sausage(!!)-related shock to his reader.
Japanese has a similar strategy, marking each word with its role, but it does so with the use of little marker words, like "ga" (subject) and "o" (object), that come after each noun:
男 が ドーナツ を 食べる。
Otoko ga dōnatsu o taberu.
Man (subject) donut (object) eats.
I can switch around が (ga [subject]) and を (o [object]) and make the donut eat the man:
男 を ドーナツ が 食べる。
Otoko wo dōnatsu ga taberu.
Man (object) donut (subject) eats.
This is a legal, if untraditional and SOV-defying move that emphasizes the weirdness of a man being eaten by a donut (implying that it’d be perfectly normal if a donut were to eat, say, a hamburger). Japanese has less flexibility than Russian; I can move the subject and object around, but the verb must stay at the end of the sentence. There is a spectrum of strictness when it comes to order: English is relatively strict, Russian is relatively free, and you’ll find languages that land between them or to either side (Chinese tends to be stricter with word order than English,4 and the aboriginal language Warlpiri doesn’t even have a default word order like the SVO of Russian).
Yet another strategy for assigning roles involves storing role information within the verb. Here’s a pair of example sentences from a modern dialect of Aramaic, spoken in northern Iraq.
The verb kemxaz-ya-le (Saw-she-him) indicates that the girl was doing the seeing, while kemxaz-e-la (saw-he-her) indicates the opposite.
Clearly, there are a lot of ways that grammar can operate when you jump from language to language, but fundamentally, it's all doing the exact same thing: it's acting as a stage director, assigning roles to each character in a story, and telling them what to do. It's the arbiter of who does what.
The better you can hold on to that idea of grammar as a stage director, the less frightening a new grammatical system will become. This is a concept that takes time to get used to, but eventually, it becomes second nature. You'll be able to look at a complex sentence like “I’m going to give my girlfriend some chocolates before our date next week” and see the chunks of information within that sentence, along with their associated roles or functions: (I - subject, chocolates - object, her - target of giving, etc. ).
Mastering a new grammatical system then becomes a matter of learning how your new language assigns the same roles to tell the same story:
Ich werde nächste Woche vor unserem Rendezvous meiner Freundin Schokoladen geben.
I will next week before our date my girlfriend chocolates give.
One way to get more comfortable with this idea is to really understand your native language. So let's take some sentences in English and look at them very closely.
Some basic anatomy of the English language -
And how to use it to understand other languages
Grammar isn't always simple, but sometimes people like to add an unnecessary and inaccurate air of mysticism to it – "You would need to be born in Japan to truly understand what the Japanese mean with this particular grammatical construction!" Ugh.
Alternatively, you may run into grammar books that casually talk about cases, registers and verb modes without really explaining what they mean. So I want to walk you through a language that you probably know pretty well – English – and show you all the grammar you're already familiar with. We’ll start with the following sentence:
John gives his girlfriend his father’s weed-whacker.
Noun Cases: The Who, The What, The To Whom, The Whose, The Where, etc.
"Cases" are just a fancy way of saying "Roles for nouns in a sentence." Our example sentence contains four of the most common cases you’ll find:
Nominative (John): Related to our word name, this case names the main player of our sentence, the subject. This tends to be the ‘basic’ form of the word, found in the dictionary. English has lost many of its case markers, but we still have them in our pronouns: you can say he did something, but not him did something. I, he, she, we and they are all exclusively in the nominative case. Our current second person pronoun, you, has eaten up four older pronouns in our language: thou, thee, ye and you. Of these, thou (singular) and ye (plural) were both in the nominative case.
Accusative (weed-whacker): Often referred to as a ‘direct object’, our weed-whacker is the target of the verb, gives, since John has to give his girlfriend something. In this case, it’s the passive player in our story, but that doesn’t have to be the case. In John fears his weed-whacker and John scares his weed-whacker, there’s quite a bit of variance in the activeness of our lawn care instrument. In the end, while these cases do define traditional roles (Nominative: Active subject, Accusative: Passive object), the individual verbs are always king, and they make demands. The words dine and devour are fairly similar, but you can’t dine a hamburger or simply devour without an object to be devoured. Cases assign roles to the nouns in the sentence; it’s the verbs who direct the play.
Dative (his girlfriend): The Dative case was known in Latin as the case for giving, and it’s common in situations involving 3 players (I sent you a can of tuna, you can cry me a river, etc.) English often uses prepositions to mark nouns for this role (I gave it to him, I made it for him). Over the course of history, we have lost most of our dative case pronouns, and with them, we have lost a clear sense of the dative case. By the 14th century, the Old English dative hwone and the accusative hwān had merged into a combined dative/accusative pronoun whom, just as hine and him merged into him, hīe and hire into her, and hit and him into it. This process continues today, as the nominative pronoun who slowly devours whom.
Genitive (his father’s): Genitive is a special case that sticks nouns together. In our example, it’s used for ownership, tying weed-whacker to his father, but there are many other reasons you might want to connect nouns. English often uses the preposition of: Heat of the sun, a loaf of bread, day of doom, etc. Our possessive construction - ’s - is one of the last remnants of a genitive case marker in English, though you can still see traces of it in some compound words like doomsday.
Nouns can play all sorts of roles, but these four tend to be the most common. Proto-Indo-European (PIE), the great-great grandfather of English, Latin, German, Russian and Hindi, had eight cases, currently covered in English by prepositions. In the PIE language, you could indicate that a given noun was a location (He did it in his backyard), an instrument (a weed-whacker made by hand), an origin (it came out of the garage) or an addressee (Samantha, John gave me a weed whacker), just by changing the endings of the nouns: Samanthap, John gave me a handot weed whacker garagen backyardi.
Proto-Indo-European (PIE) was spoken 6 millennia ago in regions of Eastern Europe and Western Asia. 3 billion people currently speak languages descended from PIE, making it the largest language family on earth. Its existence was first postulated in 1786, when British philologist Sir William Jones noticed strange similarities between Sanskrit, ancient Greek and Latin. He proposed that they came from a common source, and that source has since become known as PIE. Nearly 450 modern languages have been connected to PIE, and as such, linguists have been able to reconstruct it in detail, even down to the majority of its sounds. For a language that had no writing system and hasn’t been heard in millennia, the reconstruction of PIE is one of the most impressive accomplishments of historical linguistics.
The next time someone tries to impress you with facts like "Hungarian is the hardest language in the world; it has 18 cases!", feel free to brush them off. English has 18 cases too. We just use prepositions and context to convey them, instead of using suffixes to do it, like Hungarian does.
Pronouns, Formality and Register
“Personal pronouns and adjectives are a fruitful nuisance in this language, and should have been left out. For instance, the same sound, SIE, means YOU, and it means SHE, and it means HER, and it means IT, and it means THEY, and it means THEM. Think of the ragged poverty of a language which has to make one word do the work of six – and a poor little weak thing of only three letters at that. But mainly, think of the exasperation of never knowing which of these meanings the speaker is trying to convey. This explains why, whenever a person says SIE to me, I generally try to kill him, if a stranger.”
-Mark Twain, The Awful German Language
Pronouns, our familiar friends I, he, she, they, you, etc., act as stand-ins for nouns we’ve already mentioned, and are a common feature in most languages. They help you quickly nail down the participants in your conversations, allowing you to talk about a certain John and Sarah as ‘they’ instead of constantly going back to ‘John and Sarah did this and John and Sarah did that’. As mentioned above, English pronouns come in a few cases (and even genders!), and that’s a standard pattern; gender and case help identify precisely which participant from your previous sentence is playing a role in your new one. You’ll find that many languages have a great deal more pronouns than English. Our ‘we’ is handily broken up into four pronouns in Cherokee: ‘you and me’, ‘someone else and me’, ‘several other people and me’, and ’you, me, and one or more other people’. This would make ‘We should get dinner’ much less vague in all sorts of contexts.
Learning Pronouns without Translations
Learning pronouns without the use of English involves finding some decent pictures that show a person pointing to another person, pointing to himself, etc. I had a bunch of pronoun images commissioned for this purpose over here.
One pronoun-related nuance that English has lost is the concept of grammatical formality. Once upon a time, when we still had thou, thee, ye and you, we used the singular thou and thee to speak to a family member or a friend, and the plural ye and you to speak to a stranger or a boss. Now we can’t be quite as passive-aggressive and insulting as we once could: “Well of course ye mayeth take my sandwich from the work fridge”. Many languages use special pronouns to mark relationships between people, which demands a certain awareness of social rank, age and familiarity that English speakers just don’t need to think about.
Overarching changes in pronouns, grammar, word choice and pronunciation combine to form the concept of register - how your language changes depending on your relationship with the people you’re talking with. Register is what makes us say “I’m gunna get ice cream” to our friends and “I will send you project requirements by four o’clock tomorrow” in work correspondence. Some languages are quite demanding in this respect: Korean has an involved system of required honorific terms and verb endings covering six different levels of social interaction, between peers, different family members, subordinates, etc. The speaker’s identity, too, can come into play. To quote Sioux activist Russel Means, “the odd thing about [the movie Dances with Wolves] is, they had a woman teaching the actors the Lakota language. But Lakota has a male-gendered language and a female-gendered language. Some of the Indians and Kevin Costner were speaking in the feminine way. When I went to see it with a bunch of Lakota guys, we were laughing.” Learning which contexts demand which register is as much a cultural question as one of grammar.
Verb Tense and Aspect (Time)
English's grammar tends to be simple on many fronts compared to other languages, but we blow many languages out of the water when it comes to time. We are insanely detailed when it comes to the exact relationships of several actions, and whether they were simultaneous, completed, almost finished before being interrupted, etc. Someone can have been walking to the store for five minutes when his friend (whom he had called just five minutes before, and who had just had her birthday the previous day) calls and asks him what he might be doing the following day for lunch.
If you tease apart English’s verb forms, you’ll find that we indicate time in two different ways. Precisely when something occurred determines a verb’s tense (I saw it, I see it, I’ll see it), but I can talk about the same event in quite a few different ways: I walked (and now I’m done), I was walking (when something else happened), I used to walk (but now I drive a Segway), I’ve walked (and I’m done with that now), I had been walking (for five minutes when you called). Our sense of the verb’s progression - whether it’s completed, interrupted, never going to happen again, repeated, planned, etc. determines it’s aspect.
This gives me an excuse to talk about African American Vernacular English, or Ebonics. AAVE has a special aspect (in the present tense) that standard English is lacking, known as the habitual be form, used for describing regular, habitual actions. In one study, African American and Caucasian kids were shown a picture of Sesame Street characters Elmo and Cookie Monster. In the picture, Elmo is eating cookies and Cookie Monster is cookieless, sick in bed. The researcher asked each kid “Who is eating cookies?” and all kids pointed to Elmo. Then she asked, “Who be eating cookies?” African American kids pointed to Cookie Monster, as in “Sure, Elmo is eating cookies now, but Cookie Monster be eating cookies in general.”
Tense can be divided in all manner of ways, from the familiar past/present/future of English to the six tenses of Kalaw Lagaw Ya, a language spoken in the islands off of Queensland, Australia that distinguishes between the remote past (Once upon a time…), the recent past (I went to France last year), the today past (I ate breakfast this morning), the present (I’m eating lunch now), the today future (Dinner’s going to be delicious) and the remote future (My next birthday is going to be awesome).
A basic tenet of linguistics is that you can always convey the same thoughts in any language. English can express the same fine time distinctions as Kalaw Lagaw Ya; we just rely on context and precise time indications to do it. You can usually assume that “dinner’s going to be delicious” is about tonight’s dinner, unless we’ve already talked about a dinner date we arranged for tomorrow. In your language, you’re going to find a mixture of explicit time distinctions, expressed through verbs and other functional words (I did it yesterday, I just started) to further nail down when something happened. For a language that’s more precise in its grammar than English in tense or aspect, you will become hyper aware of time. It’s one of the odder consequences of picking up a second language - you learn to start paying attention to things you never really thought about before.
Verb Mode (Sureness, Necessity, Possibility)
There are a number of ways you can alter a verb in English in order to detail how much information you have, how sure you are about it, and the likelihood of the thing you’re saying being true or possible. These are subtle distinctions, for which we either use modal verbs “She might be in the shower” or simply add an extra verb, e.g. to think: “I think she’s in the shower”. In both cases, you’ve expressed your not-so-sureness about the situation. This is modality. Many languages have colorful and elaborate ways of expressing modality, as in this Yiddish idiom: “Az di bobe volt gehat beytsim volt zi geven mayn zeyde” (If my grandmother had balls, she'd be my grandfather), where the verb mode makes it clear that you know that your bobe doesn’t have beytsim, even if you’re talking about them. Mode can tell the listener about your attitude towards what you’re saying; many languages have verb forms that will indicate the source of your information (hearsay, confirmed facts, etc) and how much you believe it, letting you say something like “John is a good guy,” while clearly conveying that “John thinks John is a good guy, but we all know that he’s a jerk.”
Tense, aspect and mode will frequently intertwine, as they do in English. ‘Will’ can provide many different mixtures, from the simple future tense “I will eat a cookie” to the present tense, obligatory mode, “You will get me a cookie. (Right. Now.)” Different languages will mix these in different degrees and combinations.
___ am a shoe.
English has a few remnants left of verb-noun agreement, which is why you know that the proper word to fill in the blank above is "I," even if the context is odd. We’ve lost what used to be a fairly developed agreement system, where as late as the 18th century, we were still saying ‘I have’, ‘thou hast’, and ‘he hath’. What’s the point? Redundancy: languages are full of it. Even if I delete mst f th vwls n a sntnce, you can still reconstruct it; for English, studies estimate that slightly more than half of our letters are redundant, and they expect even more than that level of redundancy in our sounds. Since we both slur our words and hear imperfectly, redundancy helps us reconstruct what our conversation partner meant to say, even if failed to say completely. In some languages, you’ll find so much verb-noun agreement that certain pronouns become unnecessary. Italian is one of these, where ‘Io sono una scarpa’ (I am a shoe) and ‘Sono una scarpa’ (Am a shoe) are both perfectly legal sentences - the first one emphasizes the I, as in ‘Whatever you may be, I am a shoe’.
Function Words: Dummy verbs, Prepositions, Articles, and Conjunctions
What does “does” mean?
There are a number of words in English that can be lumped together under the broad term “function words.” These words don’t mean much of anything on their own, but they help form the basic silhouette of a sentence, immediately establishing whether it is a question, the roles of the nouns, etc. While “I waiting tell you I trouble” may be roughly understandable, “I have been waiting to tell you that I am in trouble” is much more comprehensible.
You’ll find that the most common words in any language will be its function words; the top ten list in English includes eight of them: the, be, of, and, a, to, in, and have (the other two are the pronouns he and it). These are the raw fabric of the language, within which meaningful words are embedded. And to be fair, prepositions like “in” do convey some concrete meaning - “in” tends to be a reference to the inner side of something (“It’s in the refrigerator!”), but there’s no special reason why we do things in November and on Monday. We use certain words in certain places because we have agreed to do so, and as such, the meaning of an individual function word - like “does,” above - is much less important than its traditional location in a sentence.
As with verb-noun agreement, these words are another source of redundancy that helps speakers and listeners understand each other, and every language will have a smattering of these words in nearly every sentence, to help indicate what sort of thought you’re trying to convey (like our “does” question above).
And with that, we're through with the biggest sticking points in grammar. The next time you run into a tricky passage in your grammar book that casually tosses off the term "Verb Mode" without sufficient explanation, come back here and review what that means in English, so you can realize that it's something you already know about, rather than some completely new way of conceiving of the world.
If I missed something important, let me know in the comments! While I don't necessarily want this article to become a book on its own (those books have already been written, and are excellent), I would like this to cover the terms and concepts that most people struggle with.