Grammar for Real People, Part 3: Phrase Functions and Parse · 2:24am Feb 25th, 2016
Part 1: Magic and Science
Part 2: Sentences, Phrases, and Clauses
Part 3: You Are Here!
Part 4-1: Modifier Confusion
Part 4-2: Specifier Logic
Part 4-3: Heavy Complements
Part 4-4: Positive/Negative Unity (The Truth Behind “Double-Negatives")
Part 4-5: Scripts and Sleepy Phrases (The Truth Behind “Passive Voice”)
Part 4-6: Script Details (Coming Soon)
TL;DR section down at the bottom, though you might want to read through in detail after.
Phrase Functions
So now that we know what makes a sentence and the difference between a phrase and a clause, let's look at the things that can make them up.
These are called Phrase Functions, and there are seven in all; four basic ones and three sort of auxiliaries. We've already talked about a few, so let's recap the important parts. (read: who cares about recursion seriously we ain't got time fo dat)
Specifier
This is the phrase-function version of the determiner, or the "pointing" part of a clause or basic sentence. This is actually Phrase Function #2, but we'll get to #1 after the recap. The order doesn't matter to you, anyway.
This part of the sentence or phrase specifies what the subject is, or what we're talking about. It can be just a determiner word (D word) in a simple phrase, or it can also have the subject noun itself, and anything like adjectives or other modifiers to that noun. Sometimes it can even be invisible, or at least implied or "borrowed" from context or previous sentences. In normal sentences, the specifier comes before the head, which makes sense because the head is what the subject "does."
(The technical definition of a Specifier is "the first function before the head that isn't a modifier."
"In a complete sentence, the specifier is like exposition for the sentence's story."
Head
Phrase Function #4, if you're counting. I'm explaining things as they come up, instead of the order they come packaged in. This is the phrase-function version of the tense indicator, or the "asserting" part of a clause or basic sentence.
The head is always a single word, and is the core word that the rest of the phrase hinges on. Whatever type of word the head is is the type of phrase the head belongs to (i.e. verb head=verb phrase, etc.). If you had to strip away words from a sentence while preserving meaning, the head would be the last to go.
"In a complete sentence, the head is the most basic thing you're saying about the specifier."
Now let's introduce some new ones.
Modifier
(Phrase Function #1)
If the head of a phrase is Leonardo, then the modifier is Michelangelo. It's the most modular of the modules. It would work nearly or just as well if it were moved somewhere else in the sentence, and if you were to leave it out completely the basic meaning of the sentence as a whole wouldn't be lost. In fact, you can have more than one in the same phrase. Modifiers are like the miscellaneous section of the sentence, and can be all sorts of things, from exclamations to quotes to framing comments. If something is marked by a comma, it's a modifier. Once you see a couple examples you'll be able to spot a modifier anywhere.
One more thing: items in a list are all modifiers, and the "and's" or other conjunctions between them (visible or not) are phrase heads. These phrases don't always have specifiers or complements, though they can't be complete sentences.
"In a complete sentence, the modifier works equally well at the beginning or the end."
Complement
The last of the four basic functions, and #3 in the list. Think of the complement as the head's sidekick, or interpreter, or herald, or whatever. It stays next to the head, and separating the two is a no-no (aside from a couple exceptions that I myself don't fully understand yet). The normal place for a complement is after the head, but it can --in some circumstances-- come before it. Some complements are doubled, but we'll get to that later. We said that a modifier could be deleted without altering the basic meaning of the phrase, but that's not true of the complement. The nature of the complement is to change the meaning of the head, or to add meaning to it. Once you've found a head, its complement is really easy to find. I do it by asking the head as a question: if the head is "can," I ask, "Can what?" and the part of the phrase that answers that question is the complement.
(Asking "Who can?" or "What can?" instead leads you to the specifier. Modifier is whatever's left.)
"In a complete sentence, the complement gives meaning to the head."
These four functions make up almost every sentence and most larger phrases, sometimes without a modifier. The other three don't add anything new, and really aren't important to you. A PREDICATE is the combination of a head and its complement, an ASSERTION is the combination of a specifier and the head, and RECURSION, like we've covered, is when a specific phrase function is filled by another phrase.
Parse
Now let's quickly learn some notation so we can explain things better.
"Parse" in linguistics means to break a text down to show its functions. There are two levels, basic and detail.
Simply put,
basic parse
is
just putting the functions on separate lines.
< (Detail) parse> DOES (...)
... {come (with (...) ) }
... < {special} notation> for (...)
... <each> function
, and {can (get (...) ) }
... {really} complicated.
Detail parse isn't always that bad, don't worry. It's just parentheses, and each type indicates a different function. The ellipses are just so you don't have too much stuff on one line. Every line beginning with ellipses fits into the ellipses in the line above.
Here's a key to detail parse, and some examples:
{MODIFIER}
<SPECIFIER>
HEAD
(COMPLEMENT)
"See, he does care."
{See}, <he> does (care).
"Your story is that you didn't know, huh?"
<Your story> is (that you did not know) ,{huh}?
<Your story> is (that (you did not know) ) ,{huh}?
< <Your> story> is (that ( <you> did (not know) ) ) ,{huh}?
Usually the first level of detail parse is as far as you have to go.
You might have noticed the big capital DOES in the first detail parse example, followed by "come," where a normal sentence probably would have said "comes" in place of both. This is the invisible word thing I said we would talk about back in the first blog. Grammatical structure is full of hidden connections like this. "Does come" works in the place of "comes," but it might be better to say that "comes" is a short version of "does come." When you split it out, you find that one of these words ("does") is actually the head of the sentence, and "come" is a crucial part of the complement phrase that follows it.
So for a detail parse, and oftentimes for basic parse as well, we include every invisible word we can think of, and mark them as such with all-caps. If it helps you, you can also italicize the re-conjugated verb next to the newly-revealed "invisible" tense word, like so: DOES come, but I can't really speak for how official that is in true linguistic circles.
I find this stuff fun, but I'm just weird like that. No one expects you to do this for your writing, but trying it once or twice so you understand it really helps to start seeing phrase functions at first glance, which in turn will help you avoid and/or catch wonky writing.
TL;DR
Phrase Functions
Modifier is optional for the sentence's basic meaning, is what commas exist to mark, and works just as well at the beginning or the end. There can be more than one in a phrase.
"You know, I do think so."
Specifier is the first thing before the head that's not a modifier. It's the "pointing" part of the sentence.
"You know, I do think so."
Head is the core/hinge word that everything else revolves around.
"You know, I do think so."
Complement almost always comes right after the head, and separating them is almost always bad. It alters and shapes the meaning of the head.
"You know, I do think so."
Predicate is Head+Complement, Assertion is Specifier+Head, and Recursion is when phrase functions are phrases themselves.
Parse
Two kinds of examination:
basic parse
is
just functions on separate lines,
and {...}
... <detail parse> DOES (...)
... {include and capitalize words that are usually invisible}, AND {use special notation to show function type}, and {allow for closer examination}.
{I must say},<this> is (very interesting stuff).
Looking forward to more.
3780096 Alternatively,
Because what the sentence is actually saying is that "you must say this."
The next one (which was posted prematurely the other day) should be finished and up today, and actually useful in your writing.
But yeah, this part is fun.
Checking for understanding here.
The sentence "I played all of the guitars just now." would have a detailed parse that looked like the following:
<I> DID (play all of the guitars) {just now}.
It is strange to think that every sentence contains either "to be," "can," or a visible or invisible "to do," but I can see the utility of such a notion. Invisible clause-causers seem to be a thing, as well, given your above addition of "THAT."
I can see how applicable these rules are, as something claiming to be in English that didn't follow these rules of sentence, phrase, and clause construction would be, to one degree or another, incomprehensible, but how important are the other rules? What other rules even are a thing in this particular ruleset? Here's hoping this is elucidated in your next post.
Thanks again for this series. It's been educational.
3782250 Yep, that would be correct!
All sorts of things can be invisible. If you think about it, you can usually find invisible stuff in almost any sentence, whole phrases, even.
The things we've talked about so far don't just apply to English, they're the foundations of all language, even sign languages. Think about caveman-speak, like "Me want sandwich." All it is is the basic phrase functions, each reduced to one word. It's just how we communicate.
These aren't really rules, though, just like laws of physics aren't exactly "rules." They exist whether we want to follow them or not, and you can't break them so much as break yourself against them.
There is some difference between languages in how these functions apply, like romantic languages having tense indicators built into their verbs, so the verb itself is usually the head word instead of a separate tense indicator like English: "manger" (French) vs. "to eat" (English). At least, I assume so. They could have some weird invisible stuff going on, too, I suppose.
3785662
Great. Now I'm going to have to figure out how these rules apply to German, especially when modals (such as wollen, meaning "to will") and weil (means "because") constructions come into play and shunt what would have been the sentence's main verb to the end of the sentence where normally the verb has to come as the second word in the sentence.
<I> DO (like sugar), {because <it> is (so sweet)}.
becomes
<Ich> mag (Zucker), {weil <er> (so süss) ist}.
A direct, word-for-word translation of the above would be:
I like sugar, because he so sweet is.
Looks like detailed parse still functions, even when the syntax order is all mangled by a language's weird, idiosyncratic rules.
3785687 I don't know anything about German, but it might be that the order is just different. I'm pretty sure Eastern languages are different that way, too. Or there could be some shunting going on. English does shunt stuff around some times, we just haven't gotten to that yet (not sure if we will or not).