LINGUIST List 13.2407

Sat Sep 21 2002

Disc: Do We Need a Replacement for *(...)

Editor for this issue: Karen Milligan <karenlinguistlist.org>


Directory

  1. Laurence Horn, Re: 13.2369, Disc: New: Do We Need a Replacement for *(...)?
  2. Dan Everett, RE: 13.2404, Disc: Do We Need a Replacement for *(...)
  3. Nobue Mori, Re: 13.2369, Disc: New: Do We Need a Replacement for *(...)?

Message 1: Re: 13.2369, Disc: New: Do We Need a Replacement for *(...)?

Date: Sat, 21 Sep 2002 12:50:53 -0400
From: Laurence Horn <laurence.hornyale.edu>
Subject: Re: 13.2369, Disc: New: Do We Need a Replacement for *(...)?

Joost Kremers writes,
>
>Being a syntactician, I regularly encounter the markers "*(...)" and
>"(*...)" to indicate obligatory inclusion and exclusion, respectively.
>For example:
>
>1) I see *(the) car
>2) I see the (*a) car
>
>(1) indicates that the phrase is ungrammatical if "the" is left out,
>whereas (2) indicates that the phrase is ungrammatical if "a" is put
>in.
>
>...
>We can use the plus character "+" for this marker. It is an obvious
>choice for several reasons: it is present on a standard keyboard, it
>can be used even in ASCII texts (e.g. in e-mails) and it obviously has
>some positive association.
>
>With this plus sign we can write the following:
>
>1) I see (+the) car
>2) I see the (*a) car
>
>Personally, I find this a lot easier to read. The intended meanings
>are immediately obvious.


I would vastly prefer to retain the current system. The point is not 
that (1) is grammatical with "the", it's that it's ungrammatical 
without it, which a grammaticality symbol does not express. Further, 
there already is a widely used "grammatical" symbol, the check mark. 
That tends to be used when one is contrasting a grammatical version 
with a previously discussed ungrammatical one; it's a bit like the 
natural sign used in music where one would expect a sharp or flat, 
I've always thought. In addition to the semantic problem mentioned 
above for your + nominee (or at least for the interpretation of it as 
'grammatical'), there's a nice symmetry to Ross's system (at least I 
believe it was J. R. (Haj) Ross who introduced the *(...) notation) 
that would be lost if we shifted to the plus. If the author is 
worried about possible misinterpretation, nothing prevents a remark 
at the beginning of the paper to the effect that (in the words I use 
in a book I published in 1989),

"X*(Y)Z signals that the string XYZ is grammatical but the string XZ 
is not, while X(*Y)Z signals the opposite".

I've certainly seen such a notational remark in many other works on 
linguistics, and I don't find the addition of one such sentence to 
place an onerous burden on the reader, even one unfamiliar with the 
convention.

best,
Larry
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: RE: 13.2404, Disc: Do We Need a Replacement for *(...)

Date: Sat, 21 Sep 2002 18:06:08 -0300
From: Dan Everett <dan_everettsil.org>
Subject: RE: 13.2404, Disc: Do We Need a Replacement for *(...)

	Tim Stowell's reply (Linguist 13.2404) to Kremer's question
about the *(...) notation shows a well-developed ability to reason
about the grammar of *-type notations. But his reply misses the deeper
problem of such notations generally. This can be seen in his
statements like: "The conventions of (*X) and *(X) --or (?X) and
?(X)-- were introduced to enable us to combine examples with different
levels of acceptability. Of course we can avoid Kremers's problem by
abandoning this habit, citing both examples in full, each marked with
the appropriate level of acceptability..."

	A 'level of acceptability' is inherently a statistical
notion. I doubt that any content could ever be attributed to such
levels aside from statistics. To say that * vs ** is a discrete
opposition, as opposed to a matter of context, or day-of-the-week, or
ability of the researcher to make herself clear, etc. is naive. Even
for our own intuitions. Any reader of LinguistList can check this
out. Take your favorite linguistics article with *, **, *(...), etc
markings and walk around checking the marks against native speaker
judgments wherever you like. There will be no universal
agreement. This is because all such markings make the 'homogeneous
speech community' idealization. I will not quarrel here with that
notion, but to translate that idealization into the real world
requires statistics, because there is no such speech community, as
Chomsky has made clear himself from the outset.

	Stowell also says that: "...one advantage to having an arsenal
of notations to convey varying degrees of absolute acceptability is
that there is a real difference between a marginally acceptable
example and a completely unacceptable one." And what is this 'real
difference'? It seems to me rather that 'completely unacceptable' is a
cover term for a quantification over speakers. Therefore it is
inappropriate to use the term in the absence of the quantification. Is
a given utterance completely unacceptable for all speakers of a
particular dialect in all circumstances? (Has anyone ever taught a
syntax class where all 'unacceptable' examples were unacceptable to
all students?) If not, what does 'completely unacceptable' mean?
Statistical studies are needed for the most part to give such
statements content. Thus symbolic representations of such statements
by * or **, etc. are merely rainchecks for the implied
quantification/measurements. We linguists use such devices mainly
because few of us have been trained in standard social science
research methodology. (The same reason that undergraduates take
psychology instead of physics to satisfy science requirements at many
universities - they don't want to do the math.)

	I do not personally plan to abandon the *, *(...), notation in
all future publications. But I do not continue to use it because I
think that it is the right way to look at data overall, but because it
is often good enough for the immediate point I want to make. It is
this relative usefulness of the notation that keeps it around, even
though it is clearly inferior to other methods more generally. Let's
try to avoid the embarrassing claim that such markings capture
anything real or true (e.g. that speakers judgments on degrees of
grammaticality are absolute). And let's realize that when we use them,
we are admitting that we haven't done the math.

Dan Everett

Professor of Phonetics and Phonology
Department of Linguistics
University of Manchester
Manchester, UK
dan.everettman.ac.uk 
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: Re: 13.2369, Disc: New: Do We Need a Replacement for *(...)?

Date: Sat, 21 Sep 2002 22:21:23 -0400 (EDT)
From: Nobue Mori <nobuewam.umd.edu>
Subject: Re: 13.2369, Disc: New: Do We Need a Replacement for *(...)?


 How would you then indicate:

*?(?the)

where it means:

*? without 'the', and
? with 'the'.

 The gramatical judgments indicated just outside
the ( ) means 'such and such judgments without the 
phrase in ( )', and those inside, 'such and such 
judgments with the phrase in( )'.

 Since two positions are necessary anyway, why not 
change the positions, then:

*?(the)?

This would mean '*? without the 'the' and ? with 'the'.
How's that?

Sincerely,


Nobue Mori
nobuewam.umd.edu
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue