One of the properties that seems to apply to human languages across the board is hierarchy. And as we’ve explored in past episodes, every level of grammar has some degree of organization to it — sounds, words, sentences, and meanings too. But hierarchy comes in many flavours, and there’s more than one way to organize what you say.
For instance, in our episode about the complexity of language, we talked about how there are a number of ways we might categorize the rules that build up our sentences. And depending on which kinds of rules a language uses, we can say that language falls into one of four levels: Type-3, Type-2, Type-1, or Type-0.
A Type-3 language uses the simplest kind of rule, which we can represent schematically like this:
S → word + PHRASE
This kind of rule says that a sentence ’S’ consists of a single word, followed by a phrase that can be filled in with additional words later on. This kind of rule is very limited, because it can only ever add words onto the end of the sentence — never the beginning, or the middle. And that poses a problem for representing a language like English, because we don’t build sentences by simply adding words onto the end. Take (1), below.
(1) The woman with a daughter kissed the man
At a minimum, this kind of sentence would have a structure that looks a little bit like this:
We know that those first five words must form a group, because we can treat them as a unit and replace them with a pronoun like “she.”
(2) She kissed the man
Yet, the Type-3 rule that we mentioned earlier can only ever generate structures that look like this:
Not only can’t it group the first five words together in the right way — separately from the phrase “kissed the man” — the whole tree is pointed in the wrong direction! This means that a language like English needs phrase structure rules, which can create the kinds of groups we need to accurately represent what’s going on.
Phrase structure rules are Type-2 rules, and they allow a language to group words together in ways that Type-3 rules can’t. Some example phrase structure rules include the ones just below.
- S → DP + VP
- DP → D + NP
- NP → N + PP
- PP → P + DP
- VP → V + DP
That first rule says a sentence consists of two phrases, which can each be built up out of other parts according the other rules. For example, the first half of a sentence is made up out of a determiner phrase (DP), which in turn is composed of one determiner (words like “the” and “a”) and one noun phrase (NP); that noun phrase can then be broken down into even more pieces (say, a noun like “woman,” followed by a prepositional phrase). Because that first rule lets you expand either the beginning or the end of the sentence, it can accurately represent what’s in (1), along with many, many other combinations.
In principle, though, a language can exist without these kinds of rules. An alien language, for instance, might only make use of Type-3 rules, rendering it technically simpler than ours, while still being useful for communication. A relatively straightforward sentence like “I see you” can be constructed without resorting to Type-2 rules; so, even if such a language were limited in the kinds of structures it uses, it could still function.
And, of course, such a hypothetical language could just as easily go in the opposite direction. With levels of complexity all the way up to Type-0, there’s a lot of room for speculation. In fact, we don’t even need to go all the way to the top to imagine something truly alien!
In addition to the idea that human languages can be described using phrase structure rules, we’ve also argued that we need to go a little bit further, claiming that we actually see evidence for the need to use Type-1 rules. In particular, we’ve argued that on top of the rules which build sentences, we need rules that transform them. That is, we have reasons to think that, once a sentence has been put together, some of its pieces get shifted about.
Questions are a good example of this notion; it seems sensible to think of the sentence in (3b) as being related to the more basic form in (3a), by way of a transformation. In this case, a transformation that moves that “what” from the end of the sentence to the beginning.
(3a) They are saying what
(3b) What are they saying?
What’s relevant here, though, is that these kinds of transformations aren’t without limits. And not just in English; they’re restricted in ways that consistently cut across all human language. So, while you might think we can use the same transformation from (3) on the sentence in (4a), our attempt ends up failing miserably.
(4a) Amy must love the girl who said what
(4b) *What must Amy love the girl who said?
That “who” seems like it’s getting in the way of moving the “what” to the front — a phenomenon that we see in both English and other languages (e.g., Swedish).
(5) *Vad frågade Jan vem som skrev?
‘What did John ask who wrote?’
But we might expect an alien language to work differently, with no such restrictions, or completely different ones. Like, maybe you can move a word around whenever you want, as long as you replace it with a special kind of ‘seat filler’ (not unlike the resumptive pronouns you find in languages like English and Yiddish). Or maybe any two words can switch places in a sentence, but only if they have the same category, or play similar roles, or bear the same prefix.
As we spelled out in the episode, even when we’re veering only a little bit outside the boundaries of language as we know it, we can concoct some pretty weird scenarios. In case aliens ever do land their ships down here on planet Earth, an open mind and some outside-the-box thinking will likely be invaluable, if not downright indispensable!
So how about it? What do you all think? Let us know below, and we’ll be happy to talk with you about what alien languages might look like. There’s a lot of interesting stuff to say, and we want to hear what interests you!
Previous Topic: Next Topic:
Nosing Around Phonetics Coming soon!