Message-Id: <m0rgjS6-00001pC@xiron.pc.helsinki.fi>
Date:         Tue, 21 Feb 1995 01:28:23 +0000
Reply-To:     ucleaar <ucleaar@UCL.AC.UK>
Sender:       Lojban list <LOJBAN%CUVMB.bitnet@FINHUTC.hut.fi>
From:         ucleaar <ucleaar@UCL.AC.UK>
Subject:      Re: ago24 & replies
To:           Veijo Vilva <veion@XIRON.PC.HELSINKI.FI>
In-Reply-To:  (Your message of Sat, 18 Feb 95 18:44:12 EST.)
Content-Length: 5265
Lines: 129

Jorge:
> > I think {lohe mlatu cu broda} means
> >    Ax x is a typical instance of cat archetype -> x is broda
> Is this what you mean?
>        lo'e mlatu cu broda
>        ijo ro fadni befi lo'i mlatu cu broda

Yes.

> If yes, then do you agree that {mi pensi lo'e mlatu} would be
> nonsense, since I couldn't possibly think about so many cats?

Not nonsense, but certainly untrue.

> I'm more inclined to think of it as:
>        lo'e mlatu cu broda
>        ijo su'o cnano befi lo'i mlatu cu broda

The relevant difference is my {ro} vs. your {suho}. But see
below for how we may come to agreement.

************************************************************************
* > > I'm not reading your mind properly, but I wonder if your idea    *
* > > might be something like the Mr Cat stuff brought up by Bob       *
* > > Chassell last year vis a vis massifiers. The Mr Cat idea is that *
* > > you do not distinguish one cat from another - all cats count     *
* > > as the same cat. It turned out that this is not what massifiers  *
* > > do, but it sounds a bit like what you're saying about {lohe}.    *
* > Yes, I thought about it too. What would you say to that idea?      *
*                                                                      *
* I would support this, for three reasons. First, it is easier to      *
* define that notions involving archetypes, which seem dependent       *
* on particular theories of cognition or what have you. Second,        *
* it provides a way of saying something that Bob Chassell and          *
* Mark Shoulson have mistakenly (it eventually transpired) thought     *
* {loi} does. Third, it is a nice idea that cannot be expressed        *
* by existing devices, whereas you have shown above that our           *
* present understandings of {lohe}'s meaning can be paraphrased.       *
************************************************************************

> In some sense, that's what an average is, it blurs all the members
> into one.

Blurring all members into one, yes. Average - not really. Say there
are 2 brodas. One is 5' tall, and the other is 7' tall. In this
case, I think (a) is true and (b) is false [with {lohe} meaning
undifferentiated individuals].

 (a) lohe broda is 5' tall and 7' tall
 (b) lohe broda is 6' tall

> Some questions that I'm not sure how to answer. Are these true?
>        lo'e mlatu cu mlatu

With lohe as "generic", yes.
With lohe as "undifferentiated", yes.

>        lo'e mlatu cu cmima lo'i mlatu

Is that {lohi ro mlatu} - the set of all cats? Yes, for either
meaning of {lohe}.

>        lo'e mlatu du lo mlatu

Yes, for either meaning of {lohe}.

> > > Did you mean {le du'u makau du _la'e_ lu lo'e mlatu li'u}?
> > Yes.
> > > That's equivalent, I think, to {le du'u makau du lo'e mlatu}?
> > > But I don't see what distinction you are making.
> > It might be equivalent to {le duhu lohe mlatu ku du makau} (I'm a bit
> > worried about quantifier scope in your version.)
> If I were feeling malicious I would ask you to rephrase that with the
> four more words that you need to get rid of the kau. Then we could be
> certain about any quantifier scope problem.

I can't find a way to it in Lojban that preserves use of {lohe}.
Here's a non-lojban rendering:

   Ex duhu: Ez x=z, Ay y is member of z <-> y is typical instance
      of cat archetype

With "undifferentiated" meaning of {lohe}:

  da zohu ... duhu da du lohe mlatu

> > I propose {klama} gets an extra tersumti, for the activity of going.
> > If you go twice (along same path), then this tersumti could be filled
> > by {re da}.
> There's {reroi} for that.

I know. I think of {roi}'s meaning in terms of re da in an implicit
event tersumti.

> > If you go in the present then this tersumti could be
> > filled by {lo cabna}.
> Again that's the tense {ca}. Are you proposing a new way to deal with
> tenses?

Yes. Even if this new way doesn't get licensed to appear syntactically,
if we decree that it's how thinks work implicitly, in the semantics,
then (a) it would afford a way to define the meaning of {roi} and
tenses, and (b) it would mean that a seduhu without explicit
indication of tense may nonetheless correspond to a duhu with specific
tense.

> > At present no selbri have these "event places". I say I want most
> > rather than all to have one, because a few, like {du} and the
> > mathematical ones are hard to construe as events - {lo nu da du de}
> > is odd.
> I guess all that take only sumti with abstract referents would give
> odd events {lo nu da valsi} is as odd as {lo nu da du de}.

Most would agree with you. I happen to work in a grammatical theory that
holds that words are actions, with times, places and agents.

> > If klama, zvati have this "event place", then {le se duhu zvati/klama}
> > has an implicit {zohe} in that place. This {zohe} could be specific,
> > in which case {le se duhu klama} is true iff the going happened at
> > a specific in-mind time.
> I think that in this respect you can think of the tense as implicit.
> So that {mi klama} is not just {mi klama zo'e zo'e zo'e zo'e} but
> {mi do'e klama zo'e zo'e zo'e zo'e}.

This is what I advocate. But it must be stipulated. It doesn't
follow from anything else logically or necessarily.

---
And