How Do We Signal What’s Important When We Talk? Information Structure


So let’s talk about information. When we’re talking to someone, we need to
be able to tell the difference between what’s important and what isn’t. We need a system that works for us, every
hour of every day — sifting out the old from the new and keeping track of the topic of
conversation, so that we know exactly what we should pay attention to. But how do we decide what we should focus
on, and what’s just irrelevant background noise? I’m Moti Lieberman, and this is The Ling
Space. A lot of work goes into building a sentence! We have to choose our words, and check if each one’s
been built up out of the right parts, in the right order. We’ve got to know how to pronounce everything
in just the right way, and how neighbouring sounds affect each other according to our
language’s phonology. And then we need to group our words, so all
the pieces fit together syntactically. And we have to do all this while making sure
our meanings fit the context and communicate what we want them to. Of course, we’ve got conversational maxims
to help us out along the way, like staying concise, truthful, and relevant. These are the rules we follow to help us cooperate
with each other, so that everything we say fits with who we’re talking to and what
we’re talking about. But what are we actually doing when we talk? Like, what are we trying to accomplish? What’s the point? Well, one way of thinking about it is that
we’re working to expand our common ground, which we’ve talked about before. The common ground is the shared knowledge
that exists between two or more speakers. It’s not everything that each of them knows,
because everyone has their secrets. It’s more like a big collection of everything
that I know you know, and everything that you know I know, and everything I know you
know I know . . . and so on. Think of it like a long list of things we’ve
agreed to agree are true. And when we have a conversation, we work towards
adding to this list, either by asking for information or offering it up for free. But we can’t dump just any old sentence
into the common ground. Not everything slots in there. For one, sentences can be separated into what
they assert and what they presuppose. In other words, that’s the content we’re
trying to add, and the assumptions it carries with it. So, if I say something like “Fusco’s helping
us with the case,” that assumes a few things, like we both know who that is, and we know
about the case I’m bringing up. But mostly I’m just offering up some new
information, and so it won’t have a hard time at all getting into our common ground. But if I say “Fusco’s stopped helping
us with the case,” I’m not just claiming he stopped, I’m taking it for granted that
at some point in the past he was helping. And if that’s not something you were already
aware of, the sentence will come as a surprise to you at best, and be impossible
to understand at worst. So, there are conditions of entry into the
common ground for our sentences; you only get in if you easily connect up with what’s
already there. We can accommodate making small assumptions,
like how you probably won’t think it’s too weird if I tell you “my dog only understands
commands in Dutch,” even if you didn’t know I already had a dog. But if I just popped in and told you “the
government’s top-secret, super intelligent surveillance system only understands Dutch,”
you’d be rightly skeptical if we hadn’t already established that such a thing exists. And it turns out that when we construct our
sentences, we’re not just following the same blueprint every single time; we’re
structuring the information in a way that maximizes its chances of making into the common
ground and building our conversational world. So, what are some of the ways that we do this? Well, the distinction between what a sentence
asserts and what it presupposes gets to the heart of how we structure information. What we care about when we’re trying to
communicate is which part of our information would be considered new, and what would be
considered given. This difference between what you expect your
conversation partner to know and not know is so important that languages have found
a bunch of ways to point it up and package information to fit the needs of the moment. If we don’t structure our sentences the
right way, things can actually get really strange. So like, if somebody asked you what Mr. Reese
went to get and you said “it was JOHN who got the bazooka,” that would be really weird. Even if you knew Mr. Reese and John were the
same person, and even though it does answer the question, the structure and intonation
just seem . . . off. So, how do we make sure that conversations
flow as smoothly as possible? One way to signal which part of a sentence
contains new information, and which is presupposed, is by our choice of determiner, which we’ve
seen a bit of already. Roughly speaking, if we’re introducing a
new idea into the conversation, we’ll tend to use the indefinite article, like “there’s
a machine that spies on you.” But if something, or someone, has already
been established earlier on, it’d be much more natural to use a demonstrative determiner
like “this” or “that”, or the definite article. You might now say, for instance, “the Machine
knows everything about you.” It doesn’t even need to be something you’ve
recently talked about: if I said “the government’s after us,” the country we were in would
make it clear enough who I meant. But one of the biggest ways we mark new information
is by using focus. In English, we mostly do this by putting prominent
stress on a word or phrase, so it stands out compared to background information. Like, if someone asked “who loves Sameen,”
a standard kind of reply would be something like “SAMANTHA loves Sameen,” where the
new material is focused, highlighting it as newer than the rest, which was already part
of the question. And we don’t just use focus for answering
questions; we can also do it to make corrections. Like, say someone thought that Nathan survived;
to adjust their understanding, you might say “FINCH survived”. Or we can stress a word to emphasize the truth
of a statement, like in “Harold DID see Grace on the computer monitor.” The common thread linking all these uses of
focus is that it’s calling attention to the fact that we’re selecting from a set
of possible alternatives. We’re saying that amongst all the individuals
that could fill the slot, this is the one that fits. And while manipulating the information structure
of a sentence usually has more to do with outside packaging than internal content, focus can even
affect the meaning, under the right circumstances. Like, let’s say John and Harold are good
guys who want to support people in trouble. Then the sentence “John and Harold always
help the VICTIM” should definitely be true, since they’re not going to go around helping
criminals. But “JOHN AND HAROLD always help the victim”
can easily turn out be false, if there are other people who also help. Now, focus can really land anywhere in a sentence,
to underscore what’s new. But, there’s also a bias in English towards
starting off with older information, establishing where things are at, and then following it
up with something that moves the discussion forward. So, take the sentence “as for Bear, he’ll
be fine,” or “as far as most people are concerned, Harold Finch is just a teacher.” Those phrases “as for” and “as far as”
pick out what’s called the topic of the sentence — what the sentence is about. And aboutness is the other big concern that we
have when we communicate. These phrases help us by setting up what’s
called a topic-comment structure, which introduces something to be discussed, and then says something
new about it. And while English uses certain phrases to
do the job, other languages are what you could call topic-prominent – like Japanese, which has
dedicated morphemes that pick out the topic. All you have to do is stick a は on something,
and you immediately know that’s the topic. Take our Bear sentence from before: it’d
be ベアーは大丈夫です。 Still other languages, like American Sign Language, tend
to move constituents to the front, past the subject. And this kind of topicalization is something
we see a bit of in English, too. So, the default word order in English is subject – verb –
object, as in “I have my doubts about Elias.” But if we were already talking about him,
and whether or not he was trustworthy, it might also make sense to set up the sentence
like “Elias, I have my doubts about.” And, in fact, we often use special word orders
to help with the flow. So, like, if we generally want to keep old
information first, and new last, we could passivize a sentence whose object had already
been mentioned, as in “the perpetrator was stopped”. Or if we really want to focus in on something,
we can use what’s called a cleft sentence, which can either start with “it” or “what.” So, while by default we might say “Zoe fixed
the problem,” if we wanted to emphasize that no one else did, we could say “it was
ZOE who fixed it.” Or if we wanted to emphasize what she did,
instead, we could say “what Zoe did was FIX THE PROBLEM,” which focuses in on the action
instead of the agent. Finally, these specialized word orders, like
passives and clefts, aren’t only useful for structuring the flow of information, they
can also help us probe the internal structures of sentences. Only words grouped together as a syntactic
chunk called a constituent can be passivized or clefted. So if we can say “it’s up on the roof
where you’ll find John,” we know that “up on the roof” must form a cohesive
part of the syntactic structure. And because we can’t say “the Man in was
seen the Suit,” we know “the Man in” doesn’t form a constituent; we can only
really say “the Man in the Suit was seen,” since that noun phrase and prepositional phrase
form a pair. Whether it’s helping hold together conversations
or exploring the structure of a language, when it comes to information, it’s definitely
a topic that’s important to focus on. So, we’ve reached the end of The Ling Space
for this week. If you followed my comments on this week’s
topic, you learned that part of the reason we have conversations is to increase the common
ground shared between speakers; that we have to structure the information in our sentences,
to make sure they connect in the right way; that we use things like intonation and word
order to help us focus in on new information, and keep track of the topic of conversation;
and that these tools give us valuable insights into how sentences are built! The Ling Space is produced by me, Moti Lieberman,
and directed by Adèle-Elise Prévost. This week’s episode was written by Stephan
Hurtubise. Our editor is Georges Coulombe, our music
is by Shane Turner, and our graphics team is atelierMUSE. We’re down in the comments below, or you
can bring the discussion back over to our website, where we’ll have some extra material
on this topic. Check us out on Tumblr, Twitter, and Facebook,
and try dropping by our store! And if you want to keep expanding your own
personal Ling Space, please subscribe. And we’ll see you in two weeks. Tikala malamu!

Leave a Reply