The semantics of group nouns, part 1

…in which I try to convince you that groups are atoms, and that semantics is serious business. Also, quantification.

Essentially, what I’m interested in (at the moment) is what we mean when we say something like The cricket team laughed maniacally. At first glance, this seems pretty much a non-issue. Even though I use a singular noun (the team), it is clear that the thing I have in mind is not some singular entity but a collection of such entities. We could say that the cricket team is really short for the cricket team members, and it is these members that are laughing maniacally.

At the same time, though, a group is more than the sum of its members. When I say John’s My Little Pony collection is huge, I’m not telling you something about individual Little Ponies, I’m telling you something about the collection itself. Maybe this still seems like a non-issue to you. There are simply two different ways to think (and talk) about groups: one emphasises the group as a single complex entity that is more than the sum of its parts, the other emphasises the parts. In this view, it is really rather silly to wonder whether group nouns should be analysed as ‘singular’ or ‘plural’ – they can be either, depending on how you think of them in a particular situation.

However, I don’t think this is a silly question at all, or I wouldn’t be a semanticist. Formal semanticists (like me) are not really interested in subtle nuances of word meanings; rather, we want to know how word meanings combine and interact with other word meanings to form the meaning of larger phrases and sentences. Take the sentence John is a huge idiot. I don’t really care what it ‘means’ to qualify as huge, or as an idiot, but I do care about the fact that the meanings of huge and idiot (whatever they are) can apparently be combined into a ‘larger’ meaning that we interpret as ‘very high on the idiocy scale’. (But note that we can’t use John is a tiny idiot to mean ‘John is very low on the idiocy scale’! Now that’s the stuff that makes a semanticist’s heart beat faster. See also my MA thesis and references therein.)

As it turns out, whether a word refers to a singular or a plural entity makes a difference in the way its meaning can combine with other meanings. I’m not talking about agreement (although we will talk about agreement in the next research post!) or about things like plurality-sensitive words that are incompatible with singular nouns (like together or each other). These are all syntactic issues in the sense that the words in sentences like *The boys is sleeping or *John likes each other wear their incompatibility on their sleeve: put them together and the result is just plain wrong (syntacticians and semanticists use an * to indicate ungrammaticality – don’t confuse it with the * used by historical linguists to indicate words or sentences in a reconstructed proto-language). What I’m talking about is the difference that referring either to a singular or to a plural entity can make to the possible interpretations a sentence may have. Let’s look at an example. First, a sentence with a clearly plural subject:

  1. The linguists are walking or cycling.

According to most people, sentence (1) is ambiguous between two interpretations (or, put differently, sentence (1) can be either true or false in a particular situation depending on how you read it). We can interpret it as equivalent to ‘the linguists are walking or the linguists are cycling’ – an interpretation that is true if all the linguists are enjoying the same kind of physical exercise, but false if part of them are doing one thing and the rest the other. But we can also interpret it as ‘each of the linguists is walking or cycling’, an interpretation that is compatible with such a ‘mixed’ situation.

Now, if we take all the word meanings in sentence (1) and put them together, we can only explain the first, non-mixed interpretation: we have a collection of linguists (the linguists), and a disjunction (… or …) that tells us that this collection either has the property of walking or the property of cycling. But all this gives us is a statement about the group of linguists as a whole, which is sufficient for the first interpretation, but not for the second one: ‘each of the linguists is walking or cycling’ is a statement about individual linguists, not about a collection of them.

The only way to derive this second interpretation is to assume that there is an additional semantic mechanism at work here that isn’t represented by any word in the sentence. This hidden mechanism is quantification. Those of you with some background in logic will probably recognise the quantifiers and ; in linguistics most determiners, like the, a, every, most, no, few and more than two, are analysed as similar functions. Essentially, what such an element does is take a set, and proceed to tell you something about the proportion of individual members of that set that have a certain property. Every syntactician sleeps tells you that every member of the set of syntacticians has the property of being asleep. No princess likes toads tells you that no member of the set of princesses has the property of liking toads. So this is exactly what we need to derive our mixed interpretation of (1): something that takes the collection of linguists and tells you something about the individual linguists in that collection, rather than about the group as a whole. If we assume that such a quantifier is hidden in the sentence in (1), we can explain how we get the interpretation that every individual linguist, rather than the group of linguists as a whole, has the property of either walking or cycling. (In the literature, this hidden quantifier is usually called the distributivity operator.)

Next, let’s look at another sentence that’s like (1) except the subject is a group noun rather than a plural:

  1. The group (of linguists) is walking or cycling.

Unlike (1), this sentence is not ambiguous between the mixed and the non-mixed interpretation: it can only mean that all linguists in the group are doing the same thing (either walking or cycling). This is pretty awesome because it tells us that groups, despite the fact that we can easily think of them as consisting of parts, are not equivalent to plurals in the formal semantics of a sentence. Why not? If group nouns could refer to sets, we would expect to be able to quantify over these sets by means of the distributivity operator, and sentence (2) would be ambiguous just like sentence (1). So the fact that this isn’t possible tells us that group nouns cannot refer to sets: whatever they refer to must be an indivisible, ‘atomic’ entity.

‘But’, you might say, ‘what about the example we started out with?’ True: in The cricket team laughed maniacally we do seem to say something about individual team members. But this isn’t really a problem. Formal properties of words and phrases alone do not determine how we interpret a sentence: there’s still the contribution of the word meanings themselves. So even if we regard The cricket team laughed maniacally as a statement about a team – something formal semantic mechanisms can derive without any hidden quantifiers, since it’s basically just what the sentence says – this doesn’t prevent us from reasoning about parts and wholes and exactly what it means to laugh maniacally. If I say I am in the garden, you will probably conclude that my left hand and my nose and my pancreas are in the garden, even though I never told you that every part of me is in the garden. Similarly, when I say The team laughed you will probably conclude that the team members laughed – not because I secretly tossed in a hidden quantifier, but because you know that laughing requires lungs and a vocal apparatus and teams do not have these but the parts that the team consists of (its human members) do.

So the intuition we had in the beginning of this post was at least partly right: conceptually, groups do have a part-whole structure, and this influences how we interpret statements about groups. But formally, groups are indivisible atoms, which influences how they can interact with other elements in the sentence.

When I told this story to some native speakers of British English, however, they weren’t terribly happy with it. The reason? In British English the ‘mixed’ interpretation of sentence (2) does exist with group noun subjects… but only if you use are instead of is.

  1. The group is walking or cycling (no mixed interpretation)
  2. The group are walking or cycling (ambiguous, just like sentence (1)!

I’ll try to solve this mystery in the next post. Stay tuned!

About hannadevries

University lecturer (in linguistics/artificial intelligence) with occasional opinions on religion & social justice-related stuff.
This entry was posted in Language & linguistics and tagged , . Bookmark the permalink.

One Response to The semantics of group nouns, part 1

  1. Pingback: Semantic intermezzo 1: ‘every’ and ‘only’ | hanna de vries

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s