Print

Print


In my first post the other day I suggested 24 possible degrees of
specificness when considered from the view point of 1) the speaker 2) the
hearer 3) a third party (if in existence).


In my second post I suggested that it might be advantageous to analyse the
indefinite/definite system of natlangs using my 24-way system.


Well I thought I should follow my own suggestion and in the last 48 hours I
have been doing exactly that for English. Namely ... I have analysed the
uses of the particles "the" and "a" in the English language. Also I have
looked into one use of the word "this" and "that". I mention "any" in
passing.


To recap ... ORIGINALLY I  said that logically you could gave the
specificity of a noun in a 24-way system (or we can say a noun can exist in
24 different situations). Namely ...


The noun in question can be either specific or non-specific to the speaker.
(2 choices)


The speaker can consider the hearer to know the noun in question
specifically, to not know the noun in question specifically, or the speaker
can be ignorant about how the listener knows the noun in question (3
choices)


A third person could exist or not exist, if they exist, the speaker can
consider this third person to know the noun in question specifically, to
not know the noun in question specifically, or the speaker can be ignorant
about how this third person knows the noun in question. (4 choices).


Actually I have had second thoughts about 24 possible degrees of
specificness and have REDUCED it to a 8-way system.


 =============================================================================================


OK ... now I am going to digress a little. In my last post I was using
terms such as 1st person specific / 2nd person non-specific and 1st person
non-specific / 2nd person non-specific . I don't think these terms are
helping at all. Algebra existed for a long time with very little progress
being made : all through the Greek Age and the Roman age and the Middle
Ages. It was only when an efficient notation was devised that people could
started to manipulate the different terms and progress was made in Algebra.
I am going to attempt to do the same for indefinite pronouns (well actually
I want to devise terms for my 24 different "situations"). Please bear with
me ... it is not that difficult.


Each term will begin with a capital S (maybe meaning "situation" or "in the
situation" ... it doesn't really matter). Then there will follow 2 or 3
values.


The first value will be "1" if the item is specific for the 1st person, "0"
if not so.


The second value  "1" if the item is specific for the 2nd person, "0" if
not so. If the speaker does not know we have an "X".


The third term is blank if no third person is involved, "1" if the item is
specific for the 3rd person, "0" if not so. If the speaker does not know we
have an "X".


So ... in the sentence "She wants to marry *the* Norwegian", we can say
*the* represents S 1 1 1 (pronounced ... es wun wun wun) ... I am trying to
spell phonetically ;-)


In the sentence "during my trip to Budapest I met *this *really nice girl",
we can say *this* represents S 1 0 (pronounced ... es wun zeero)


In the sentence "Do you know *that/the* guy that got drunk last night ?",
we can say *that/the* represents S 1 X (pronounced ... es wun eks)


[ Just to re-iterate ... in my notation a "1" means that Z can be
identified as one particular Z out of all Z that exist. A "0" means that Z
can not be identified as one particular Z out of all Z that exist. If the
"1" or "0" is in the initial slot, we are looking into the mind of the
speaker. If the "1" or "0" is in the next slot, we are looking into the
mind of the hearer. ]


OK ... we have the notation. Now let us consider the situations one by one.
I am designating every situation with letter from the alphabet. However
this is only for convenience in this thread (hopefully somebody will want
to discuss this with me).


So first off ...


(A) S 1 1 .... this one is easy. specific to both speaker and hearer, part
of the body of knowledge that they share. I consider S 1 1 to be "stable"


[ By the way, I consider the term "specific" to be exactly the same as
"definite" and the term "definite" to be exactly the same as "referential" ]


(B) S 0 0 ... this one is also easy. The item is non-specific to both
speaker and hearer. S 0 0  is "stable" also.


The next 4 situations represent a possible mismatch between the knowledge
that the first person and the third person have. Now assuming* that the two
protagonists are from the same family or tribe : that they are friends or
colleagues (the usual situation) the information would probably be shared.


[ *Also assuming that the item under discussion is important. But actually
to say this is not needed. If the item under consideration was unimportant,
nobody would be talking about it, would they? ]


(C) S 1 0 ... This one is my favourite. I consider it pretty unstable ...
humans are a gregarious lot and like to share information. In English there
are two particles used to tag a S 1 0 noun : both "a" and "this".


When physically introducing a new item to a person it is common to use
"this". For example "come and look at this colourful little insect". Well
of course "this" core meaning is/was for drawing the hearers attention to
something near the speaker. The hearer typically being further away from
the object. Right away you get connotations of ... "seen better by the
speaker" => "understood better by the speaker" => perhaps "known only by
the speaker" and hence in modern day English "this" is used for introducing
an unseen  S 1 0 object. And just as when you introduce a visible object
with "this", when you introduce a distant object with "this", the
expectation is that you are going to talk a bit about the object ... to
make it from a S  1  0  to a  S  1  1 .


If you heard from an acquaintance "during my trip to Budapest I met
*this *really
nice girl" you would expect to hear quite a bit about her. Whereas if you
heard "during my trip to Budapest I met a really nice girl" there would be
no such expectations. Maybe the latter sentence was a reply to you saying
"all hungarian girls are unfriendly to foreigners" ... "during my trip to
Budapest I met a really nice girl" is just your acquaintance rebutting your
assertion.


So I would say, an S  1  0  object, when tagged with "this", is likely to
change into a S  0  0  object. While an S  1  0  object, when tagged with
"a", is more likely to stay an S  1  0  object.  However these are just
tendencies, not rules. There is nothing wrong with introducing an S  1  0
object with "this" and then saying nothing more about it, and there is
nothing wrong with introducing an S  1  0  object with "a" and then
expanding on it.


I am going to mark S  1  0  objects tagged by "this" as "unstable". By that
I mean they are likely to change into S  1  1  objects.

I am going to mark S  1  0  objects tagged by "a" as "stable". That is they
are likely to remain S  1  0  objects.


Now you should have figured out what I meant by "stable" in situations (A)
and (B).


OK ... we have covered the first three situations. The remaining 3
situations are quite different. They all engender questions*. So instead of
the objects being tagged with "the"/"a"/"this", they are sought. They are
SOUGHT. Sought with "who"/"what"/"which".


* Well they engender questions only if they are deemed significant, if they
are considered insignificant : only part of the background of the message,
then they would continue to be referred to marked (somehow) as unknown.


(D) S 1 X ... Pretty unstable. The speaker will want to ascertain if the
hearer can identify the object that he wants to discuss. So he ask question
to that effect. If the hearer replies in the affirmitive then immediately
the situation changes :

              (S 1 X  =>  S 1 1).  If the hearer replies in the negative,
the speaker will, in all probability, enlighten him. A two stage process
leading to the same result  : (S 1 X  =>  S 1 0  =>  S 1 1)


NB .... where as in (C) the speaker will use "this" if he intends to talk
more about a certain object. In (D) the speaker might use "that" in a
question to ascertain if the situation is S 1 1  or  S 1 0 *. For example
"do you remember that girl that was really freaking out at her boyfriend
last night" **.  In this case "girl" is S 1  1  but maybe she is not that
prominent in the hearers memory. The reason for  "Do you remember that girl
that was really freaking out at her boyfriend last night" is to raise
awareness of the girl in the hearer's mind ... to make her a solid S 1  1
(as opposed to a  S  1  0.5  or  a  S  1   0.3  ..... or even  S  1  X  if
the hearers attention was wondering the night before).


*  I find it pretty neat how the usage of "that" in (D) mirrors the usage
of "this"  in (C).


** This is a good time to point out that we are not talking digital here
... more analogue ... more a spectrum of values than two discrete values.
But of course there is always an attraction in considering a situation as
either black or white ... it makes it so much more simple.


(E) S 0 1 ... Along with S 1 0  the most unstable. If you don't know
something and your mate does ... well, you will want that information.


(F) S 0 X ... Unstable. How the situation changes mirrors (D). For example
... the speaker knows that Jane had been raped a few days ago in the town.
So, of course, he knows a human male agent was involved. But he doesn't
know who. Assume he is visited by his brother from the town. The question
will be either "who raped Jane" or "do you know who raped Jane". That is
either a content question or a polarity question. In this case we have ...


"who raped Jane"   "I don't know"
   S 0 X  =>  S 0 0

"who raped Jane"   "Mad Hugo"        S 0 X => S 1 1


"do you know who raped Jane"   "Mad Hugo"     S 0 X => S 1 1

"do you know who raped Jane"   "no"  S 0 X => S 0 0

"do you know who raped Jane"   "yes"     "well who" (annoyed voice)
S 0 X => S 0 1 => S 1 1


[ Maybe the last dialogue shouldn't be included. To answer "yes" in this
situation is abnormal. ]


Actually there is not much difference between (E) and (F).



Notation Stability Comment Mode(s) of change


(A) S 1 1 stable the noun tagged with "the"

(B) S 0 0 stable the noun tagged with "a"

(C) S 1 0 stable (with "a") the noun tagged with "a"

S 1 0 unstable (with "this") the noun tagged with "this" : portends a xfer
of info. => S 1  1

(D) S 1 X unstable (on the whole)   engenders a polarity question
                       => S 1  1

=> S 0 1 =>  S 1 1


(E) S 0 1 unstable engenders a content question => S 1 1


(F) S 0 X unstable engenders a question ... if a content question ... => S
1 1 => S 0 0


engenders a question ... if a polarity question ... => S 1 1

=> S 0 0

(=> S 0 1 => S 1 1)


Earlier I proposed considering how well a third party knew the object under
discussion. My initial system was very complicated *. On second thoughts
the full system I was thinking about is untenable.


* I was going to include all of  S 1 1 1 , S 0 0 1 , S 1 0 1 , S 1 X 1 , S
0 X 1 , S 0 1 1 , S 1 1 0 , S 0 0 0 , S 1 0 0 , S 1 X 0 , S 0 X 0 , S 0 1 0
, S 1 1 X , S 0 0 X , S 1 0 X , S 1 X X , S 0 1 X . Definitely overkill.


However I would like to keep two of these ...


(G) S 0 0 1


(H) S 0 0 0


The third person I am talking about, could be anybody. If one person can
identify the object under discussion it is obviously real.


If nobody can identify the object under discussion it is not (necessarily)
real or at least not real at this point in time.


The sentece "she wants to marry a norwegian" can be given an S 0 0 0
interpretation and an  S 0 0 1  interpretation.


Side Note ... [ I find it a bit strange that English has no way to
differentiate between (G) and (H). If "she wants to marry any norwegian"
was grammatical this diffentiaion would be made. However to me it doesn't
sound grammatical* ... nearly grammatical but not quite. To me the
sentence  "she wants to marry a norwegian ... ANY norwegian" is acceptible.
I also find the negative  ("she doesn't want to marry any norwegian")
acceptible, but for some reason that I can't put my finger on, "she wants
to marry any norwegian" is not quite right. ]


So that is my analysis of English according to my system.


 =============================================================================================


At the moment I am reading INDEFINITE PRONOUNS by MARTIN HASPELMATH. An
interesting book. He defines 9 "situations" (page 64 if you have the book
handy).


I notice that what he calls (1) "specific known" = S  1  0


What he calls (2) "specific unknown" = S 0 0 1


What he calls (3) "irrealis non-specific" = S 0 0 0


Now he has (8) "comparative" and (9) "free choice ... I think I could
include one or both of these, quite easily into my system. I am not sure
what I would call them. For now I will just say S ???


I can draw a little implicational map for my system ...


        S 1  1    .........   S  1  0   .........  S  0  0  1  .......   S
0  0  0  .....  S  ???


Specificity decreases as you go to the right ======>


Now my systenm is 1 dimensional. Haspelmath's is 2 dimensional. Haspelmath
devised his map by a the hard slog of comparing how many many languages
handle indefinite pronouns. I devised my system from logic. I find it very
pleasing that they are not incompatable to each other. I would like to
expand my system tho' ... and thereby explain the whole of Haspelmath's
map. A tall order but that is what I would like to do.


If anybody reading this has an L1 other than English, or are really well
acquainted with a second language, then I would be really interested to
hear their analysis of their language within my framework. Unfortunately
grammars written about languages, even if otherwise excellent grammars tend
to skip over the subject of indefiniteness.


Looking at Samoan (from David McCann's data) it looks like "la" is the tag
for S 1  1 , "le" the tag for S 1 0  and "se" the tag for S  0  0.


I wish I know more about Lakhota. Would I be able to fit it into my
framework. (Maybe the second definite marker something like S  D  D  where
0.3 < D < 0.7 ... just a thought)




              Best Regards ... Stewart Fraser



On 11 November 2015 at 08:48, stewart fraser <[log in to unmask]> wrote:

> I am glad you like the system that I proposed.
>
>
> Actually I was not really serious about creating a conlang with this
> 24-way distinction (well for one thing it would be far too
> kitchen-sink-ish) : rather I was thinking that ... say you came across a
> natlang with a 3-way distinction ... maybe my proposed 24-way framework
> might be of use* to analyse what is actually going on in the natlang.
>
>
> Probably the main point of my post was to try and set the terminology
> straight.
>
>
> By the way, I would condider "definite" = "referential" = "specific". I
> choose to use "specific" for no major reason.
>
>
> I agree with all the points you make. The logophoric idea is very neat.
>
>
> * The 9 "situations" as Haspelmath proposed in his book "Indefinite
> Pronouns" would be another useful framework to use in analysing this 3-way
> natlang. Actually "situation" 1 and "situation" 2, which he calls "specific
> known" and "specific unknown" correspond to what I call 1st person
> specific / 2nd person non-specific and 1st person non-specific / 2nd
> person non-specific .
>
> On 10 November 2015 at 22:53, Alex Fink <[log in to unmask]> wrote:
>
>> On Tue, 10 Nov 2015 03:41:32 +0000, stewart fraser <[log in to unmask]>
>> wrote:
>>
>> >It would be possible to make a really comprehensive set of particles in a
>> >conlang to cover all possibilities.
>> >
>> >You could give 1st person the value of specific or non-specific. (2
>> choices)
>> >
>> >You could give 2nd person the value specific, non-specific or unknown. (3
>> >choices)
>> >
>> >You could give 3nd person the value "no 3rd person involved", specific,
>> >non-specific or unknown. (4 choices)
>> >
>> >So you could make up a conlang with 24 particles ( i.e. 2 x 3 x 4 ) to
>> >cover all possibilities.
>>
>> This is a nice system, in both the old and new senses of "nice" (but not
>> the obsolete sense!).  It's the sort of thing I might even quasi-steal for
>> a conlang one day, though I'd probably make some alterations.  For one, the
>> more grammaticalised a system is, the less likely it is to have a
>> noncommittal option.  Your "unknown" values are not quite noncommittal
>> options but I suspect the distinction you actually describe between 'I know
>> you don't know which' and 'I don't know whether you know which' would be
>> actually solidly makable in few cases, so it'd have more or less the same
>> effect.  For two, instead of having third person orthogonal to first and
>> second, it could be fun and less overbuilt to have it come in through the
>> back door of logophor.  That is, there'd be four articles contrasting in
>> known vs. unknown to speaker and known vs. unknown to hearer; but in
>> contexts where a third person's viewpoint is being taken the "speaker"
>> system would actually be used to mark said third person's perspective (and
>> this change could come with a morphological marking, or no).  If one really
>> wanted to indicate knownness to all three parties one could use a double
>> article, maybe along the lines of '[unknown-to-s-unknown-to-h me.DAT]
>> known-to-s-unknown-to-h N' where the inner tighter-bound article shows the
>> logophoric interpretation and the dative pronoun overcomes this
>> interpretation for the outer one.  For three, I wonder if the
>> unknown-to-speaker-known-to-hearer article should be identical to a
>> question word with the force of 'which'.
>>
>> Alex
>>
>
>