*old response on imperatives and the FAQ

[Logical Language Group <lojbab@ACCESS.DIGEX.NET>]

>From LOJBAN%CUVMB.BITNET@UBVM.CC.BUFFALO.EDU Tue Nov 28 21:41:01 1995
>Date:         Tue, 28 Nov 1995 19:33:50 -0700
>From: Chris Bogart <cbogart@QUETZAL.COM>
>Subject:      Special meaning of V-initial
>To: Bob LeChevalier <lojbab@ACCESS.DIGEX.NET>
>
>>And:
>>> > I agree. I would also prefer that V-initial not be singled out like
>>> > that.
>>> Do we know why it is? A relic of some ancient Brownian predilection?
>
>Jorge:
>>Something to do with the Loglan imperative, I think.
>
>I also remember seeing an argument related to relative phrases something
>like this:  in a {poi broda} phrase it's likely that you'll want x1 to
>be {ke'a} and to explicitly state x2.  If V-initial weren't special, and
>if syntax within a poi were consistent with sentence-level syntax, then
>you'd have to explicitly use {fe} or {zo'e} or {ke'a} to get to the x2.
>
>For example, now we say {le nanmu poi prami mi} and the x1 of {prami} is
>elided, and we can assume it's {ke'a}, which here equals {le nanmu}.
>Without this special treatment of V-initial, we'd have to say {le nanmu
>poi prami ke'a mi} or {le nanmu poi ke'a prami mi} or {le nanmu poi ke'a
>mi prami}.  So:  it saves 2 syllables in what's arguably the most common
>way of using {poi}.  May or may not be worth it, depending on how you
>value word order flexibility vs. brevity.
>
>BTW should I stick this along with the historical explanation in the
>FAQ?  I do think I've heard the question before.
>                     ____
> Chris Bogart        \  /  http://www.quetzal.com
> Boulder, CO          \/   cbogart@quetzal.com


Yes, put it in the FAQ.

The main answer is indeed that Loglan uses missing x1 as the imperative.
There was a debate in 1979 in favor of missing x1 being the observative
but JCB was not convinced and he had no easy way to do an imperative.

Since I came up with "ko", we had the option.

The observative, the relative clause explanation, and the special place
of x1 in general in the Lojban sentence (as the referent of "le broda")
all justify x1 being normally before the selbri.  The latter can be
explained by comparing with the selbri form:  broda be da bei de.  In
this form we HAVE to assume some deafult place numbering, and the most
useful one is the one with x1 omitted, since x1 is subsumed in the "le"
when we convert the selbri into a description sumti.

To be consistnet with a V initial form, you would always have to say
broda be fe da bei de.

(Note that this latter also is used to justify the continuation of
numbering after a fa-tag, rather than going back and filling in missing
gaps.  If we did not continue numbering, then it would have to be
"broda be fe da bei fi de" or de would turn out to be x1.)

So to put it simply, the entire syntax of the language as a built in
bias against unmarked verb initial forms.  To allow them and be
consistent and our constructions, we would have to make SOV form highly
marked.  Maximizing UNMARKED word order flexibility has never been a
priority in the language.  Indeed, in JCB's language I think it is
ALWAYS marked when you avoid SOV order.  He created clumsy "sumti
rotator" constructions for SOV and VOS order and other devices far more
highly marked than the Lojban norms (he doesn't have zo'e either).

lojbab