Ode to RRG
In Role and Reference Grammar, the predicate semantics
Are first among the theory’s theoretical mechanics;
The model’s more than that of course: a rich and radiant prism:
Macroroles, the AUH and the linking algorithm.
’Cause RRG’s a theory which is rich and round and real;
It’s syntax’ best-kept secret but a massive function deal.
Chomsky, Goldberg, Jackendoff, these names need no explainin’!
But who tops off the list? Professor Robert D. Van Valin!
So here’s a little ditty on the theme of aspect lexical
(Please forgive that little bit of syntax asymmetrical;
It’s just to get the scansion), a.k.a. Aktionsart
Which is where the Role and Reference Grammar textbooks tend to start.
(Let’s not forget of course, friends, that the aspect labelled ‘lexical’
Is quite a different story from the aspect termed ‘grammatical’;
And that’s another poem which I’m penning at the mo
But there’s much I have to read about and lots that I don’t know.)
Yes, Vendler (1967)’s where the model starts:1
Accomplishments, activities, achievements and then states2
But RRG adapts it up and adds in two more classes:
First: active accomplishments; and second: semelfactives.
That’s the first addition to the Vendler set of four.
But RRG is bold and brave and adds a little more:
For each of these six predicates has something more to give:
Each is paired up with a corresponding causative.
Let’s take a quick example to illustrate what’s what.
Consider, please, this sentence (1): ‘The air balloon just popped.’
So that is an ‘Achievement’, a punctual change of state.
Now let’s turn to the causative if that’s OK. Yes? Great!
(But first let’s demonstrate that (1) is not a causative;
In fact, it’s what is sometimes called an unaccusative:
A verb that is intransitive whose subject is a patient
And not (as with unergatives) an actor or an agent.
Now the corresponding sentence which depicts a causative
Increases as to valency, and so becomes a transitive:
‘The bird just popped the air balloon’ (let’s call that sentence (2))
Just think of it like this: it’s like Picard says, ‘Make it so.’)
Now we could take a detour and consider all the tricks ’n’
Challenges with causatives, read up on R. M. Dixon;
The forms: it could be lexical and sometimes periphrastic;
It’s often morphological; the causative’s fantastic!
Let’s not forget the terminology one needs to know:
The arguments are often labelled S or A or O;
And degrees of causativity: ‘My captor let me go’
Is not the same in force to things like ‘Santa made it snow.’
But let’s not take that detour; we’re here for RRG
And how it builds on Vendler’s seminal typology
We’ve got a dozen classes: 6 types, each one in twos
And we’ve slowly started walking in Van Valin’s holy shoes.
Let’s do another type now; I propose Activities.
They’re dynamic and unbounded eventualities
Like ‘march’ and ‘walk’ and ‘roll’ (intransitive) and ‘swim’ and ‘think’
And ‘run’ and ‘write’ and ‘read’ and ‘jog’ and ‘snow’ and ‘drive’ and ‘drink’.
‘Are there any tests?’ I hear you ask, ‘for these 12 classes?’
Is helium the lightest of the so-called noble gases!?
In Van Valin’s ’05 book (p.39) some tests are given.
Seven tests in fact; it’s one big diagnostic heaven!
Which brings us to the formalism. Could anything be worse
Than talking ’bout the formalism in this silly little verse?
I’ll have a go in honour of Van Valin’s great advances
Even if I have to do it in these four-line stanzas.
Bare predicates are utilised to indicate a state:
know′ (x, y); dead′ (x); some others: see′, be-in′, and wait′.
The (x, y) stuff is arguments, expressing valency
It may look esoteric but it’s no big mystery.
Activities are formalised with operator do’
Then predicate and arguments (so ‘x’ (plus ‘y’, if two)):
‘eat’ is: do′ (x, [eat′ (x, (y))]); do′ (x, [cry′ (x)]) is ‘cry’
Why is do′ there? It just means ‘activity’, that’s why.
And on we go with operators for the other four:
Achievement: e.g. INGR popped′ (x) (we saw this before)
Where INGR means ‘ingressive’. Semelfactives, they use ‘SEML’:
The formalism captures the semantic fundamentals.
That’s enough. I think I’ll leave Accomplishments for you
(’N’ their friend: Active Accomplishments): read from p.42
Of RVV (’05); you’ll see it all in black and white
And even check that what I’ve penned is (largely—vaguely) right.
The Causative! Let’s not forget. It’s α CAUSE(s) β
The schema’s pretty neat so far but this bit’s even neater!
So we’ve covered all the types now, one-by-one and line-by-line,
But many verbs require these different subparts to combine.
So let’s take an example (3): ‘The doggy scared the boy’
This is Causative so, α,-β-CAUSE ahoy!
(dog, ∅)] CAUSE [feel′
])]. Wow, what a string!3
I had to write this poem to understand the flamin’ thing!
There’s much more to be written about predicate semantics
They’re always up to naughty tricks and cheeky little antics.
And there are other frameworks, if you wanna load your plate:
Comrie (’76) or Moens and Steedman (’88).
And RRG itself is far, far richer than these classes:
I’ve but read a fraction—but burnt through three pairs of glasses!.
Syntax templates, PSA, and juncture, focus, linking.
If you want something to think about, RRG’ll get you thinking!
So that’s it for this poem—but not for RRG;
It’s growing ever stronger in its second century:
Linguistics, linguists, heed it well and hold it in your soul
And reference it always as it plays its grammar role.
In fact old Vendler kicked this off a whole decade before
In nineteen fifty-seven, which to my unending awe
Was two whole years before he even got his PhD
In 1959, at Harvard, in philosophy!
Anyone who knows this stuff is likely to object
‘States comes first in Vendler’s list, and that, my friend’s a fact!’
‘I know and I agree, of course,’ is all I can reply;
It’s simply to allow the line (unlike this line) to rhyme.
I’ve read some RRG; I hope this kind of comes across
But I don’t know how Van Valin would prefer us to pronounce
The zero or null symbol in the do′(x, ∅) string;
I myself say ‘such that’ but it could be anything.