Re: Raising and Equi-verbs: a birds eye overview

On Monday, April 5, 2004, at 09:08 PM, taliesin the storyteller wrote:

> * Ray Brown said on 2004-04-05 20:55:33 +0200
>> Trask doesn't appear to mention 'equi verbs',
>
> He calls them control verbs (I just looked it up). Equi might be a bit
> old-fashioned.
..and not English - 'tis the Latin for "horses"   ;)

But thanks - yes, I've now found 'control verbs' in Trask (also, I see,
known as 'catenative verbs').

it was also becoming apparent reading replies on the list last evening
that I misunderstood Trask's definition of 'raising verbs' and that some
of my examples from modern Greek were of control verbs rather raising
verbs.

I've followed the terms up in Crystal as well. I just want to see if got
things straight. If I haven't, I would appreciate correction from the
professional linguists on this list.

1. RAISING VERBS
I guess when Trask wrote: "A lexical verb or predicate which typically
appears in a syntactic
structure in which its surface subject is logically or semantically the
subject of its complement clause" he meant by implication, '..and not of
the surface verb", i.e. in "Lisa seems to be happy", Lisa is the semantic
subject of "be happy" and not of "seems" for can equally well say:
It seems (that) Lisa is happy
_or_
Lisa is seemingly happy/ Lisa is apparently happy

That is, if I've understood it aright, the sentence "Lisa is happy" stands
at a deeper level (according to transformational grammar) as the subject
of a sentence whose VP is "seems". Thus in subject-raising Lisa is 'raised'
  to become the subject of "seems" and "is happy" is transformed to an
infinitive as the complement of "seems".  So in Latin "Lisa laeta est" +
"uidetur" (it seems) --> Lisa laeta esse uidetur.

Crystal also gives an example of object-raising where when the sentence
"John is honest" is the object of "he believes", "John" is raised to
become object of "believes" and "is" becomes an infinitive complement to
the verb, thus: "He believes John to be honest".

So in Latin, "Ioannes honestus est" + "credit" --> Ioannem credit honestum
esse - the good ol' Latin accusative-and-infintive! But whereas CL was
stuck with the acc+infin. we can, and more often do, in English keep the
object sentence as object with no raising, thus:
He believes (that) John is honest.

Indeed, with most verbs, unlike Latin, the so-called object-raising is not
an option; thus while in CL "Gaius laetus est" (G. is happy) + "dicit" (he
says) must become "Gaium laetum esse dicit", we do not in English say *"he
says Gaius to be happy".

2. CONTROL VERBS (aka 'catenative verbs', aka 'equi verbs')
The logical or semantic subject of the infinitive clause is not raised to
become the subject of the main verb but is under the _control_ of the
subject of the main verb. According to Trask, these divide into
subject-control verbs and object-control verbs.

As an example of a subject-control verb, Trask gives "promise", e.g. He
promises to visit me., i.e. "he promises" + "he (will) visit me" - te two
subjects are the same.

But to awkward Latin treats the verb as _object_, i.e. 'raises' the
subject of 'will visit' to become the object of "promise" - but we mustn't
say 'raise' because "he" is also the logical subject of "promise". OK -
but Latin definitely treats 'promise' as an _object-control verb_ (if we
use control-verb terminology), thus:
"promittit" + "me visitabit" --> se me visiturum esse promittit
The good ol' acc.+infin. again!

In English, tho not CL, we may, as with raising verbs, simply retain the
second sentence as an object clause, thus:
He promises (that) he'll visit me.

As an example of object-control Trask gives "persuade", since the object
of persuade is the logical subject of the infinitive, e.g. I persuaded him
to go <-- I persuaded him + He went.

Altho the object & infinitive construction is more common for "persuade"
in English, a noun clause is possible as object, thus: "I persuaded him
(that) he (should) go".

In fact, in Latin we _cannot_ use the acc.+infin. with 'persuadere' and
_must_ have a full clause because 'persuadere' is not a transitive verb
and cannot have a direct object. Our English object is the _indiect
object_ of 'persuadere', thus the only permitted CL equivalent is:
    ei   persuasi    ut   ueniret.
(to-him I-persuaded that he-go)
---------------------------------------------
A FEW OBSERVATIONS

By referring to 'surface structure' in his definition, Trask is by
implicitly relating 'raising verbs' to transformational grammar theory;
Crystal is more explicit and says of 'raising': "A type of rule recognized
in some models of transformational grammar".

Both Trask and Crystal quite explicitly relate 'control verbs' to
Government Binding theory of grammar which both state to be derived from
transformational grammar theories.

Therefore, it seems to me, the validity of these terms must be dependent
upon the validity of the transformational grammar theory and, in
particular, on the validity of GB theory.

The distinction between control-verbs and raising-verbs does not, in any
case, seem to me to be so clear cut as one or two have suggested. I have
shown that _both_ types of verb (if, for the sake of argument, we accept
the distinction) often map to the same accusative-and-infinitive
construction in Latin. I think, however, I would find it difficult to
convince all Latin scholars that there is a distinction between 'raised
accusative-and-infinitives' and 'controlled accusative-and-infinitives' -
and, indeed, I am far from persuaded that there is any such useful
distinction.

On Monday, April 5, 2004, at 03:48 AM, Herman Miller wrote:
[snip]
> Okay, raising verbs make sense, but I would've expected them to be some
> odd peculiarity of English (or Germanic languages in general). Is it
> really the case that all languages have them?
I don't know - but what is clear they are not the same in all languages.
For example 'must' is a raising verb in English, but in the modern Greek &
Welsh equivalents, it ain't:
cf.
MG.       prepei    na   po
     it-is-necessary that I-go = I must go

  W. [mae]         rhaid   i   fi fynd
     [there's] a-necessity for me to-go = I must go

Also, interestingly, while English has subject-raising "I must go", Latin
has _object-raising":
me ire opportet (the good ol' acc.+infin. again!)

But even more commonly the Latin equivalent is the gerund* as grammatical
subject of "to be" with the 'logical subject expressed by that dative case
(cf. Welsh 'i fi'), thus:
mihi eundum est.

*or, according to some, the impersonal neuter of the gerundive. Indeed,
where we have a transitive verb with "must", our object becomes the
grammatical subject in Latin and the gerundive agrees with that subject, e.
g.
You must send those letters -->
illae epistolae tibi    mittendae     sunt
those letters   for-you to-be-written are  = you must write those letters.
cf.
MG. prepei na grapseis  afta  ta  grammata
			  you-write those the letters
and -
  W: rhaid i chi sgrifennu'r llythyrau'na
                 to-write the letters there
[Note: the...there = that/those - literary Welsh has: yr llytherau hynny
"the letters those"]

Presumably, according to the transformationists, the forms "illae episolae
tibi scribenda sunt", "prepei na grapseis afta ta grammata", "rhaid i chi
sgrifennu'r llythyrau'na" and "you must write those letters"  are all
surface generations from an underlying form such as:
((you write those letters) is needful)

In the same mail, Herman also wrote:
> But I don't get what "equi-verbs" are supposed to be, unless it's
> "something that looks like a raising verb but isn't".
He may well have a point   ;)

Ray
===============================================
http://home.freeuk.com/ray.brown
ray.brown@freeuk.com    (home)
raymond.brown@kingston-college.ac.uk   (work)
===============================================
"A mind which thinks at its own expense will always
interfere with language."         J.G. Hamann, 1760

Reply