LINGUIST List 7.1524

Mon Oct 28 1996

Disc: Natural language

Editor for this issue: Ann Dizdar <>


  1. Miguel Carrasquer Vidal, RE: 7.1513, Disc: Natural language
  2. Waruno Mahdi, Disc: Natural language
  3. benji wald, Re: 7.1513, Disc: Natural language

Message 1: RE: 7.1513, Disc: Natural language

Date: Sun, 27 Oct 1996 02:19:40 PST
From: Miguel Carrasquer Vidal <>
Subject: RE: 7.1513, Disc: Natural language

>Date: Thu, 24 Oct 1996 14:17:49 +0300
>From: (Jouko Lindstedt)
>Subject: Re: 7.1490, Disc: Natural language
>> Are there any rules that
>> are not found in natural languages -- or, maybe more difficult to say,
>> "impossible in NLs"
>In an earlier contribution, Ivan Derzhanski in fact suggested that
>there may be such a rule in Esperanto, namely "the assignment of
>nominative case by prepositions in Esperanto" -- instead of the
>accusative which is the only other (surface) case in Esperanto.

Had this parenthetical remark been made about any natural
language, I would not have blinked. It may not be my cup of
tea, but I can read it with a straight face. In this case,
however, a mental image was conjured up of Zamenhof planting
allatives, partitives and instrumentals in the depths of 

Nonsense, of course, but did anybody experience the same 

- -----------------------------------
Miguel Carrasquer Vidal
- -----------------------------------
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Disc: Natural language

Date: Sun, 27 Oct 1996 19:58:54 +0100
From: Waruno Mahdi <>
Subject: Disc: Natural language
Having been busy with other things, I have not been following this
discussion as closely as I would have liked to, and perhaps what
I have to add here is already deja vu. But just in case...

I'd say that there is a principle or fundamental difference between
Esperanto and any natural language, in that the latter has a long
(practically unfathomably deep) history and prehistory of perpetual
change, and that the present state is just another point in this
process. That natural languages have numerous "exceptions", missing
in Esperanto (as noted, I understand, by Benji Wald), is perhaps not
the only consequence of this circumstance.

Natural languages have social and professional jargons or "dialects"
(the most ancient of the latter probably being shamanic secret language).
These jargons (to keep this term for lack of anything better) are
important in identifying the speaker as belonging within the group,
and require a steady rate of renewal to maintain their exclusivity,
as knowledge if some features seep out. Of special significance seems
to be youthese, an age-group jargon which possibly plays an important
role in language change in general. Young people use this jargon
to separate themselves from the "adult world", and are "bilingual"
in that they use different verbal signalization when talking to
peers and when talking to others (particularly adults).

There have been many observation on languages from various parts
of the world, indicating that sound changes occur abruptly, and
that, instead of a gradual series of intermediate articulation, one
finds alternation of what is practically the old and the new articulations
in the in-between generation. We thus have here the workings of "sound
law" ("historical phonology") being observable within the lifetime span
of one generation.

Another consequence of language change is that, within one synchronic
plane, some expressions are archaic, whereas others are felt as neologisms.
Some such expressions get "updated" quite often (cf. in English "blow one's
top", "fly off the handle", "go ballistic" etc.).

Esperanto seems (I'm not much of a specialist here, so please correct
me if wrong) not to have social and professional jargons, spontaneously
arising sound shifts operating within an overviewably recent or current
period of development, an historically founded archaic style, a constantly
rejuvenating youthese, periodical replenishment of "in" expressions.
I think that the moment a community should take Esperanto as its "first
language", the latter would rapidly lose its "regularity" and become no
different from any other natural language. It might be interesting to
hear how the Masoretic canons of Hebrew are faring since this started
being used as natural language in Israel. (Latin in the Vatican,
Old Church Slavonic in Russia, Sanskrit in India remain reserved for
religious use only. Real-life Arabic dialects differ considerably
from the Classical Arabic of the Qoran, the same way as the Prakritic
dialects diverged from Sanskrit).

The indicated particularities of artificial languages like Esperanto
seem mostly also to be true for machine language and animal language.
In machine languages (which undergo periodic development either because
of more sophisticated implementations, or to cope with commercial
competition), archaisms usually either become incompatible or persist
as normal ("legal") expressions, i.e. there is normally no archaic "style"
as a quality perceivable by the machine. In animal languages of species
with dialect formation (e.g. some bird species) a changed set of signals
represents the same invariable set of meanings to be conveyed. "archaic"
signals do not persist, but become "foreign" and thus unintelligible.

Regards, Waruno

- ---------------------------------------------------------------------
Waruno Mahdi tel: +49 30 8413 5408
Faradayweg 4-6 fax: +49 30 8413 3155
14195 Berlin email:
Germany WWW:
- ---------------------------------------------------------------------
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: Re: 7.1513, Disc: Natural language

Date: Sun, 27 Oct 1996 15:51:54 PST
From: benji wald <bwaldHUMnet.UCLA.EDU>
Subject: Re: 7.1513, Disc: Natural language
Jouko makes an interesting observation concerning Esperanto:

>... an apparent asymmetry is that the subordinating temporal
>conjunction 'before' is "antau ol" (lit. "before than"), but 'after'
>is "post kiam" (lit. "after when"). Though "post ol" is used at times,
>it is definitely a rare form. But notice that we do not really know
>whether this is simply an inconsistency, or whether there are some
>cognitive reasons for this distinction between the two conjunctions to
>persist (for instance, the sometimes different factual status of the
>situations referred to in the two types of embedded clauses).

My thought was that "earlier than" might be the motivation for the "before"
expression. This leads to questions about "early" and various other
lexical items in Esperanto (or the source language/s or system of logic
which inspired the inventor in this case). It would seem that the lexical
item is simply IE as in Latin ante, which may lead away from lexical
motivation to some kind of syntactic motivation. Would "post ol" be
inspired by "later than", cf. "aft-er"?

Meanwhile, the "factual status" suggestion is obscure to me ("belief
status" I would call it). My first thought was that "after" was the
culprit because of the possibility of varying degrees of certainty about
the "future" (i.e., some time after a currently observable time), but
"before" seems to have more transparent variable behavior for belief
status... so is that it? For example, "water the plants before they die"
is idiomatic English for "so that they d/won't die", but an anti-pragmatic
factual interpretation is also possible, "even though they're gonna die
anyway", i.e. "so that they'll die wet", cf. "I hope she changes her will
before she dies" (if she changes it after she dies there may be some legal
problems vs. if she doesn't change it she'll die?)

Gallows humor aside, the observation is interesting and begs an
explanation, leading to consideration of whether such a consciously imposed
construction is "natural" (i.e., possible in an unconsciously evolved

Jouko further notes:

>In an earlier contribution, Ivan Derzhanski in fact suggested that
>there may be such a rule in Esperanto, namely "the assignment of
>nominative case by prepositions in Esperanto" -- instead of the
>accusative which is the only other (surface) case in Esperanto. After
>prepositions, the accusative marker -n is only used to show direction,
>e.g. "en la urbo" = 'in the town', but "en la urbon" = 'into the
>town'. Apparently some theories of Universal Grammar predict that the
>complement of a preposition cannot be in the nominative. There are two
>ways of approaching this problem:
>When children acquiring Esperanto from one or both of their parents
>learn to use the accusative ending for direct objects (some time after
>their third birthday, I assume), we could investigate whether they
>tend to overgeneralize it after all prepositions as well. I do not
>know what the result would be. But the problem is that the child is
>practically always simultaneously acquiring another language as well;
>if this language does not have surface case, it may inhibit the
>"natural" overgeneralization in Esperanto; if it does have surface
>case, what looks like an overgeneralization may be in fact
>interference. So there might be difficulties in interpreting the data,
>but the investigation would still be worthwhile.
>Another interesting approach is to question the validity of the
>standard grammatical descriptions of Esperanto -- as linguists are
>ready to do with other languages. Notice that the "accusative" ending
>is also used after adverbs to show direction, e.g. "hejme" = 'at
>home', "hejmen" = '(to) home', and "tie" = 'there', "tien" =
>'thither'. So it might just not be a case marker at all in a
>typological sense, though its use as a marker of the direct object has
>made the name "accusative" traditional for it.>

To begin with, as Jouko implies, the labels "nominative" and "accusative"
show IE inspiration, just as the antipathy to extensive case systems by the
creator of Esperanto shows concession to Western IE languages, whose
speaskers find "case" (as learned in classical declensional paradigms)
tedious, as if (and maybe it's true?) inflectional languages are more
cumersome than more analytical languages. Of course, the whole thing about
"accusative" to express "place to which" (and "time during which") is
classic IE. An economical analysis, again as Jouko implies, might simply
see Esperanto as having an unmarked (= nominative) "case" and a marked
case, one that is only different from a preposition in its syntactic
properties. Is the possible unnaturalness based on the assumption that
"accusative" marks "objects" in a "natural" language, whether objects of
verbs or prepositions? The "postpositional" analysis of the Esperanto
"accusative" seems like a way out of the prediction that children learning
Esperanto as a first language would have to overgeneralise the accusative
to prepositions during learning.

Finally, Jouko notes:

>Incidentally, in the discussion so far creologists have been
>absent. Do they endorse the characterization that pidgins are
>"non-natural" and creoles are "natural"; and what are then "extended

Again, that's a good point, and empirical evidence is murky in the same way
that Jouko had observed for native acquisition of Esperanto, i.e., that
first generation creole speakers in observed situations actually acquire to
some extent more than one first language, the "pidgin" (extended or not)
and the first language/s of the parents. This has created all kinds of
chicken-and-egg problems for resolving controversies about the "bioprogram"
and innateness hypotheis in relation to pidgin and creole development.
Thus, for example, Bickerton's 70s studies of Hawaiian (English) pidgin and
creole assigns certain features of the pidgin (of native Japanese and
Filipino speakers) which he wants to argue are actually creole creations,
to *borrowings* from the creole (children of the pidgin speakers) back to
the pidgin. Conversely, the first generation Hawaiian creole speakers do
tend to have some competence in their parents' languages -- which allows
features that are theoretically associated with bilingualism but not
creolisation to be assigned to the bilingualism not to the creole formation
(the contribution of empirical data to resolving problems of theories
which makes different predictions for bilingual and creole phenomena are
frustrated here.)

Tok Pisin, much more than Hawaiian pidgin, is a good example of an
"extended pidgin" (as well as a creole); similarly various forms of West
African Pidgin English (related to Krio, a bona fide and longstanding
creole with many first and second language speakers). Here the same
problems have arisen in resolving empirical data with respect to cumulative
development of a pidgin so that it is grammatically indistinguishable from
a creole vs. the input of creole speakers into the pidgin -- and, also,
pidgin speakers back into the creole.

To my knowledge, situations which could resolve the issue of whether there
are linguistic correlates to the distinction between "creole" and "extended
pidgin" have not been observable, and are only assumed from speculations
about the past. These speculations, of course, were theoretically
motivated to the extent that they pertained to interpretation of linguistic
data (but also to the precise social context), and therefore have led to
circularity when they have been imposed on actual observable situations.

All this may be quite apart from my terminological query about whether
second languages are counted as "natural" languages ("natural" =
"native"?), and the answer may have to do with convenience in a specific
context of discussion, e.g., "languages" used by machines vs. languages
spoken by people, rather than some deeper or more generalised use of the
term. -- Benji
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue