Grex Language Conference

View Responses

Grex Language Item 58: Germane Gerunds
Entered by rcurl on Fri Mar 18 07:48:49 UTC 1994:

"coloring book" is a book to be colored, not a book that colors
"cutting edge" is an edge that cuts, not an edge to be cut
 What is the rule?

16 responses total.

#1 of 16 by other on Fri Mar 18 09:49:28 1994:

They are different cases.  In the former, the book is the object of the
verb color, while in the latter, the edge is the tool used in the cutting, not
the object which is cut.

#2 of 16 by remmers on Fri Mar 18 11:37:12 1994:

Hmm, I don't know that they really are different cases.  "Cutting edge"
could also be interpreted "edge used for cutting" rather than "edge that
cuts"; this would be the same as "book used for coloring".

Other instances:  curling iron, walking stick, dining hall, reading room,
looking glass, training table, shooting range, conferencing system.  All of
these can be parsed as

                <noun> used for <gerund>

"Edges don't cut people; people cut people," as the saying goes.

#3 of 16 by rcurl on Fri Mar 18 15:09:22 1994:

Falling rock.  rock used for falling? Nooooooo..  And (as pointed out
earlier) the conventional us is "conference room", not "conferencing
room", but that's a different issue. Do we have more contrary examples?
Burning Bush, boiling water, walking fern, talking head. These parse as

               <noun> that is <gerund>

So, how is the difference made clear? Is "cooking oil" oil that is used
for cooking, or that pot of oil on the stove that is cooking (like the
baking pie).

#4 of 16 by srw on Fri Mar 18 18:18:47 1994:

Re #1, I wouldn't say that in "coloring book" that "book" was the subject of
the verb "color". I believe "coloring" is an adjective in this case.

Re #3 "cooking oil" - It's clearly both. In the form created by putting
"ing" on the end of a verb is quite ambiguous in English. It could be one
of at least these cases: 

(1) the present progressive tense of the verb (water is falling)
(2) an adjective indicating its action        (falling water)
(3) as adjective indicating its use           (drinking water)
(4) as a noun (gerund)                        (falling is dangerous)

(there may be others)
btw: I believe only case 4 is a "gerund", as gerunds are nouns.

Note that these sentences are ambiguous:
(a) Drinking water is dangerous.
"Drinking" is either case 3 or 4, yielding two different meanings.
In case 3 water is the subject - in case 4 drinking is the subject.

(b) Do you see the cooking oil?
"cooking" is an adjective, for sure, but is it case 2 or 3. Who can tell?

As far as there being a rule to tell case 2 from case 3, I don't think
there is one. Ambiguity is resolved by context. That's why computers will
have to learn how to think before they will be able to translate natural
languages.

#5 of 16 by davel on Fri Mar 18 20:43:41 1994:

I think Steve (srw) is pretty much right.  There are lots of other
syntactically-ambiguous constructions not involving verbal participles,
of course.  One famous one involves things like:
  John is easy to please   (it is easy to please John)  vs
  John is eager to please  (which refers directly to John in a different way).

The way a lot of pedants object to "hopefully" is kind of similar.  They
allege that something like "Hopefully, I'll be able to go home on time
tonight" can only mean that I will be able to go home on time in a hopeful
manner.  But a little thought will provide a huge number of examples of
adverbs used in other ways, in ordinary English.  "Logically, you would
expect ..." doesn't mean that someone would have an expectation in a
logical manner, for example.

There's only so much you can actually extract from grammatical form;
a *lot* of syntactic forms are ambiguous in resolution, depending on
what is being said as well as the parts of speech used to express it.

#6 of 16 by albaugh on Fri Mar 18 21:37:21 1994:

Further, I don't believe that context is something that can be reasoned out.
It is something known from memorization.  Therefore, computer translators
must have an auxiliary store of word patterns with associated, fixed meanings,
and cannot get by merely by analyzing syntax & grammar.

#7 of 16 by davel on Fri Mar 18 22:03:55 1994:

Yeah.  Semantics is not just syntax plus single-word glosses.  "Word patterns"
is a good name for it.

#8 of 16 by gerund on Sat Mar 19 03:45:21 1994:

People *are* talking about me!  :-)

#9 of 16 by srw on Sat Mar 19 05:41:25 1994:

OK, I guess I missed one. Case 4 and Gerald are both gerunds.
Sorry, Gerald.

#10 of 16 by gerund on Sat Mar 19 07:08:51 1994:

Perfectly ok.

#11 of 16 by rcurl on Sat Mar 19 07:48:02 1994:

Gerunds are not *only* nouns, though I may have used the term over
broadly. My dictionary says that the gerund "retains certain
characteristics of the verb, such as the ability to take an object
or an adverbial modifier (Ex. Playing golf is his only excercise.)."

#12 of 16 by gerund on Sat Mar 19 08:28:24 1994:

A Gerund is the child of a Noun and a Verb.

#13 of 16 by srw on Sat Mar 19 12:54:08 1994:

Re: #11, I'm baffled. My dictionary says a gerund is a verbal noun.

#14 of 16 by davel on Sat Mar 19 13:56:34 1994:

I think a better example might be "playing fast is easier than playing
slowly".  "Slowly" can't normally modify a noun, but can modify a gerund.
But Steve's point that non-noun participles aren't gerunds is correct.

#15 of 16 by carl on Sat Mar 19 14:41:53 1994:

Isn't walking stick an oxymoron?  How can it do both at once?

#16 of 16 by omni on Sat Mar 19 23:55:36 1994:

 There are walking sticks in nature (insects) and canes which are also
walking sticks. I guess it has something to do with context.

Response not possible - You must register and login before posting.