On Tue, Jan 02, 2007 at 01:08:53AM +0100, Lars Finsen wrote:

> >My least favorite part of the book is the last chapter.  The author
> >attempts to show, in a broad sketch, how language as we know it would
> >have developed naturally and inevitably from what he calls the "me  
> >Tarzan" stage to the fully subordinating structure that we have today.
> This makes me curious. I'd like a very brief outline, please. I know  
> some people have screwed-up ideas about the origin of language. Is he  
> one of them? Or is it you? Or is it me?

Ok, I will attempt it...

His first premise: that there existed a stage where language consisted
of 2 or 3 word sentences, where the words were simple invariable words
which probably did not pattern as nouns and verbs, although they would
naturally have referred to objects and actions due to our experience of
the world.  He does allow a basic world order of SOV (not sure how he
squares this with "no nouns or verbs"!).  There are no adjectives,
prepositions or other frippery.

He starts at this stage because his claim is that existing, observed
tendencies of language change can work on this stage to create modern
language, but he believes that we cannot know what forces brought speech
*to* the multi-word stage.

Now, to summarize the chapter...

Let me quote his initial example story, and perhaps you will see why I
found it a bit patronizing-sounding:

girl fruit pick     turn     mammoth see
girl run     tree reach     climb     mammoth tree shake
girl yell yell     father run     spear throw
mammoth roar     fall
father stone take     meat cut     girl give
girl eat finish     sleep

He states about this that "speakers of any language would be able to 
follow it without any problem, as long as they understood the meaning
of each word" because it "does not rely on any rules peculiar to [...]
the grammar of any other particular language", since "the words are
strung together according to a few natural principles, which are rooted
in the deepest levels of our cognition".  That may very well be true,
but it is the sort of claim which sounds quackish.

Next he goes into more detail about his "me Tarzan" stage of the language.
He believes that we should start with only words for physical things, 
simple actions, and the closed class "this" and "that", which he justifies
despite their abstractness since they usually accompany the act of 
pointing.  He spends a couple of pages explaining that his word order 
above is based on keeping related words together (OV?), actor-first (S..), 
and chronological order.  Digresses to show ancient example of text strung 
together with "and" keeping phrases in chronological order, states that 
today's complex web of conjunctions is a recent invention.  I remember 
reading something to that effect about Native American languages that 
imported conjunctions from Spanish, possibly in Mithun...

Next he reminds us that the brain is predisposed to see structure and 
organize hierarchically, and that he is not taking a stand in the debates 
over exactly *how* this relates to language (he is Chomsky-agnostic).  
With that he gets down to business.

First topic to tackle: "this" and "that".  Better justification for 
including them as primary words: he claims that in no language have
"this" and "that" been traced back to any more concrete origin, only
alternating with "here" and "there" which are often built on "this" 
and "that" and vice versa.  Again mentions the pointing, and says that
that is why "this" and "that" are more basic than other words with
shifting references (such as "here", "there", and all pronouns).  In
fact, he then shows that pronouns develop from "this", "that" and,
where applicable, "that yonder", with language examples.

Having provided pronouns, he moves on to ways to include oblique
participants in clauses, i.e., adpositions.  His 2- and 3-word
sentence language had only actors and patients.  He states that
multiple simple clauses can collapse into one clause via a verb
turning into a pre- or post-position, and provides examples from
a number of languages.  He then declares "and" invented since it
can develop from "with".

Next he attacks adjectives.  He shows that basic words for property
objects have been known to develop from words for objects notable for
that property, i.e., red from blood, sharp from shard.  Then he 
embarks on a completely gratuitous metaphor regarding the "double life"
of property words, their "high life" as predicates (a word he does not
use to avoid scaring us) and their "low life" as "appendages" to the 
noun.  This is the sort of ridiculous paraphrase he uses throughout 
the book, BTW.

His justification for how the "property words" became "appendages" 
is rather threadbare.  He claims that "this" and "that" would have
begun to be used next to the relevant noun, as in "this stone" or
"stone, that" due to the simple need to disambiguate.  I gather that
he is trying to say that juxtaposition of two nouns could be 
interpreted by the next generation as a noun and an attribute.  Now
that I put it that way, it makes more sense.  Anyway, he says that
once you have "this" and "that" in a modifier-modified relationship
with a noun, people could begin using property words the same way
by analogy, giving us adjectives.

Then he writes "once the principle of appendagehood has been established
with one type of appendage, the flood-gates have been opened, and a 
stream of other types can gush in: plural markers, quantifiers, articles,
possessives and so on".  He then addresses each of the above in a bullet
list of paragraphs to show from what pre-existing words they could have
sprung.  Having justified all of these noun modifiers, he claims the
speakers of the language would have started using them all together
to create complex noun phrases, and would also have been open to the 
idea of stacking modifiers on modifiers, both via simple modifications
to the adjective such as comparitives and superlatives, and via 
recursion of, for example, possessives.

Then he hops over to the verb to bring it up to speed.  But first he
explains why the syntactic categories of nouns and verbs are different 
than the distinction between action words and object words, and declares 
that his language has developed abstract noun concepts (it must have 
done so in order for his property words to have developed from nouns, 
for example) and therefore now has a syntactic noun-verb distinction.  
Provides an interesting example (the derivation of the suffix -hood) 
to demonstrate abstraction from nouns.

Another interesting discussion with examples: the fact that nouns are
seen to easily give rise to verbs, sometimes simply being verbed 
wholesale, but that nominalization of verbs has been seen to arise
late and usually as a back-formation from a related verb and abstract
noun both derived from the same concrete noun.  I did not know this.
Uses Old French -age to demonstrate.

Now he addresses verbs.  Refers to previous chapters showing how 
tense markers and causative/passive forms have been seen to develop from
separate words.  Interesting demonstration of the progressive abstractification
of verb to possession to obligation to marker of likelihood, right here
in English:  "get me a beer" -> "he's got a car" -> "I've gotta go" ->
"she's gotta be there by now".  He seems to have a point.

Finally says that his language has developed the necessary preconditions
to develop subordination.  Nominalized verbs can become attributes to nouns.
He anthropomorphizes a bit and says that verbs are "used to being at the
centre of the action, with numerous participants crowding around them" and
therefore would drag other participants in along with them, creating the
first relative clause (albeit using participles).  He then hand-waves 
around how the true relative clause with finite verb form and relative-
clause marker such as "that" developed, saying that it happened before 
attested examples and must have been driven by the need for clearer, more 
explicit relative clauses.  He declares victory, as it were, and saunters 
into the epilogue.

Not so bad, really, and the examples are very good.  I just found the
father-killing-(singlehandedly!?)-the-mammoth-for-his-daughter story a
bit too precious.

Late to bed!