Steve's Site

Introduction:
So you want to learn to write in your own language better, or learn another one, and this grammar thing just keeps getting in the way? And perhaps you're a programmer or ex-programmer or wannabe-programmer, or even a webmaster -- you can find a missing semicolon in a 5-page C listing in 4 seconds flat, but you still don't know an adverb from an adjective? Well despair not, fellow geek. This one's for you.

In Steven Pinker's "The Language Instinct", he points out that all human grammars are quite similar. Though there are thousands of human languages extant, there are only 17 or so "switches" you need to set, to go from one of these grammars to the other. Of course even assuming each switch has only 2 positions, that's 128K possible grammars, but they're not all that different, really.

(If I got these numbers wrong, correct me. I'll update the entry.)

So if grammars of different languages are all so similar, we can save ourselves the pain of learning hundreds of hours of grammar lessons when we take our next language course, by simply knowing the basics. And they really are similar, the same way most programming languages have subroutines.

I'll explain human grammars here by reference to computer languages. My analogies will be based in procedural (or procedural-capable) languages like C++, Fortran, or Perl, rather than pure Object-Oriented languages. Not because you couldn't do the same with O-O terminology, but because *I* can't. In my mind, human languages are procedural languages. If you don't like that, write your own article and I'll link to it.

Chapter 1: Verbs: Operators and Functions

Note: I'm going to oversimplify drastically in the first few chapters. Pedants bear with me and I'll qualify nearly everything as the time arrives for it....

Every sentence in a human language contains a verb. A verb is the operator, or function, of that sentence. Compare:

"John kisses Marie"
kiss(John, Marie);
"Giovanni bacia Maria" (Italiano)
"Jean baise Marie" (Francais)
"Johan pussar Marie" (Svenska)

Where's the verb in each sentence? Hint: It's not John or Marie.
That's right, the verb is like the function. It makes something happen. If this were a textbook, I'd insert a bunch of "find the verb in these sentences" exercises here. But since it's not, you're on your own.

How many arguments?

Note: From this point onward, I'm going to start making generalizations that apply best to Indo-European languages, and Germanic and Romance families in particular.  Sorry about that bias, but I know next to nothing about the others, and the two languages (Hebrew and Japanese) in non-Indo-European families that I'm familiar with actually do follow these generalizations quite well.  So for those familiar with very different languages, please feel free to tell me how to be more general.

In ANSI C, a function has a prototype that details how many arguments a function takes, and what type each one is, and perhaps which ones are optional. Human languages have the same for each verb, although you'd probably have to look at a dictionary to find the spec. But in your own language, you probably already know it. In English for example, what's wrong with:

John goes Marie. (too many arguments)
John kisses. (too few arguments)

See? You already know which verbs take 2 arguments and which don't.

if you were a programmer, you could make a simplified prototype like
go( subject );
strike( subject, object);

and not be too far from the truth. In fact, if you look up your favorite verb in a dictionary, you'll see the abbreviation "v.t." or "v.i" after it.

v.t. (verb, transitive), means the verb takes two arguments, subject and object.
v.i (verb, intransitive), means the verb takes one argument, the subject.

You may also find a v.t. entry and a v.i. entry for the same verb! Now the O-O people get happy because they like polymorphism. Yes, when a verb changes number of arguments, it generally changes its meaning a bit too. Just like in Java.


Can a verb take more than 2 arguments? This answer depends on your point of view, and your mother tongue. A third optional argument, the "indirect object", may be present (even for intransitive verbs!). Many grammar teachers would be unhappy to look at it that way, but this is not their book.

In this book, we'll just say YES, the indirect object *is* one of the verb's arguments. To set it apart from the direct object, most languages require extra tag words (prepositions or postpositions) to indicate and modify indirect objects:


---------------------- Sidebar: Prepositions --------------------------------
The prepositions "a (Italian), over, at" have one main function and one optional function. The main function is to point out which noun is the indirect object. The optional function is to show HOW that indirect object is involved. If you've ever tried to learn any foreign language, you'll notice that the second function is remarkably inconsistent. There are literally hundreds of ways for an indirect object to be involved, and only a few prepositions, so don't be surprised if the next language doesn't use them all the same way. In many cases it's virtually random.

e.g.
Describe that in English. (why "in"?)
Descrivilo in Inglese (why "in"?)
Beskriva den på Svenska (why "on"?)

I think about London (why "about"?)
Io penso a Milano (why "at"?)
Jag tanker om Stockholm (why "around"?)

In a few cases where a preposition is obviously appropriate, it will likely be the same in all languages. (eg. "under the sofa, in your stomach") But usually it's peculiar to the language. Esperanto tries to solve this problem by having a general-purpose preposition with no set meaning, for use in all cases where "over, under, above, etc." don't obviously apply.
---------------------------- end of sidebar ---------------------------

Ok, so what are all those arguments in a real language?

In human languages, most verbs take three arguments -- a subject, an object, and an indirect object. Some can be left out.

Since there are thousands of different verbs, the definitions of these three terms are necessarily very vague, to cover lots of ground. But rest assured there's no chance you'll confuse one with the other once you get the idea.

Subject: The person, place, or thing DOING the verb.
Object: The person, place or thing DONE BY the verb.
Indirect Object: That other person, place or thing, usually RECEIVING or LOCATING the object of the verb, and usually tagged with a preposition (at, into, under, to, etc.).

So:
John threw rocks at Bill.
Subject: John
Object: rocks (being thrown)
IndObject: Bill (receiving rocks)

Bill threw rocks at John.
(left as an exercise for you)

I wrote letters to Marie.
Subject: I
Object: letters (being written)
IndObject: Marie (receiving letters)

The cat barfed under the sofa.
Subject: The cat
Object: vomit (left out but understood, being barfed)
IndObject: the sofa, underneath.

Note that the order Subject, Verb, Object is NOT true in all languages, or even always true in English. Japanese uses Subject, Object, Verb. Don't even get me started on German. But you can safely guess which is which by asking "Who's doing it"?, "What is it being done to"? and "Who receives the results"?

In many languages (including Japanese, Russian, Latin, and Esperanto) the subject, object, and indirect object are marked using special suffixes or small tags. These languages just make it easier to parse the sentence, that's all! Other languages like English and French have to make do with strict word-order rules instead.

Now make up some simple sentences, and find the verb, subject, object and indirect object, until it feels obvious.

Come back tomorrow and we'll talk about NOUNS, a.k.a. constants.