This thread is getting a bit complicated, but I think it needs to stay that way as for now. So, in this message I'm going to offer more precise semantic definitions for FL1 and FL2, and compare them to paaS. Here's for accusative ambiguity in FL1:
>> ((assault (topic)
>> (((boy (accusative (topic)
>> ((((interior (house (accusative (topic)
>> (((girl (nominative (topic)
[Daniel:] > The "topic" is one thing but I take it there are two distinct instances of
"accusative"? But how would I conclude that without already knowing the
The English phrase was "A girl assaults (a boy in the house)" where both a boy and the house are part of the object; i.e. in the accusative.
With the grammar S ‒> aS, where a is in S, more plausible object sentences are:
(((boy (accusative (topic)
((((house (inessive (accusative (topic)
where the house is in inessive (in-location), inessive is in accusative, and accusative is in the topic, which actually is not the topic, but a case of the topic (a topic-case). The topic can include several such cases, which can be nominative, accusative, locative etc. Like in natural languages, verbs can also be put in cases, but actually I want to change strategy here, because there's an easier way to do the job.
Namely, the grammar S ‒> aS, where *a is in a role of S*, both the boy and the house are part of the object, which is one of the roles of (i.e. under) the topic. With this definition we can ditch nominative and accusative altogether and use subject and object roles instead:
(((assault (verb (topic)
(((boy (object (topic)
((((house (interior (object (topic)
(((girl (subject (topic)
I think this is looking a lot neater, check out:
- assault is in a role of verb
- boy is in a role of object
- house is in a role of interior (in-location); and importantly, interior is in a role of the object (it follows logically that if you kick something which is included in the interior of a house, you kick the interior of that house; however the house is not a role of the object.)
- girl is in a role of subject.
- all are roles of topic. So the topic has many roles, which are syntactic roles (role1).
It follows from the formula that each recursive element is a role marker for the initial element. So you can say that there are two syntactic classes: initial element (a) and role marker (S). There's only one lexical class which is noun. Boy can be a noun, but with an added hyponym it takes the role of a role marker (see example below).
I think in my research I'm going to use a slightly different formula. From linguistics POV all FL1 sentences are endocentric compounds where A+B (a+S) denotes a special kind of B. I have to check the terminology, but you could say that A is a specification of B. This way we have as an example
farm+boy, a boy who relates to, or is specified by, the farm.
While I think it's not directly a case of ambiguity, the relationship in this example is relatively vague. To have a more precise language we can stick to the above grammar S ‒> aS, where a is in a role of S. Here one could argue that farm is not sufficient for a role of the boy. However, an unlimited number of specifiers, or intermediate roles, may be inserted, e.g.
for "boy in the farm".
Next, Logan mentioned the possibility of a translation relay, and I saw there was a potential beginning for one, which would not only involve FL1 or FL2 as opposed to English, but also PL and PAS (I think I understand the difference);
We had the meaningful English sentence "So you are in trouble because you lost your wallet."
> [Logan:] So, here's my interpretation of that sentence into predicate logic:
(∃e.∃w. lost(e) & ag(e,you) & wallet(w) & gen(you,w) & pat(e,w)) ->
The English sentence can be translated into FL2 (aSb, where b is a role of a) as:
[so [subject you] [in trouble] [reason [time past] [subject you] [object [of you] wallet] lose] are]
and into FL1 (brackets added, 'so' removed) as:
(((you) subject) existence)
(((trouble) in) existence)
(((lose) reason) existence)
(((((((you) of) wallet) object) lose) reason) existence)
((((past) time) lose) reason) existence)
(((((you) subject) lose) reason) existence)
Note that in a top-down reading it becomes a "police report" type of expression:
EXISTENCE > subject > you; inside > trouble; reason > LOSE > object > wallet > of > you; time > past; subject > you.
Where I've introduced some optional conventions: capitals to mark head items, marker (>) for a direct relation, and the elision (;) for a skip back to last head item. This is not really important though.
> Rearranging terms gets us to
> (∃e.∃w. lost(e) & ag(e,you) & wallet(w) & gen(you,w) & pat(e,w)) ->
(∃e. in_trouble(e) & pat(e,you))
> Which differs from my original intuitive formulation only in that we
have reified the event of "being in trouble" rather than simply
asserting that the referent of 'you' satisfied the constraint of being
Right, there can be several potential parses for the English sentence. I think maybe the syntax I used - same as the Stanford Parser, was more like
(∃e.∃w. lost(e) & ag(e,you) & wallet(w) & gen(you,w) & pat(e,w)) ->
(∃e. in_trouble(e) & ag(e,you))
Anyway, as for paaS, the CKY parser generates just one instance of recursion, but I tried the "brute force parser" of JFLAP which does give an endless right recursion, the parse tree being (paa (paa (paa (paa (paa)))))…
In other words paaS has one infinite branch. This is modest in comparison to FL2 which allows n items on level 2, each of which have n subitems on level 3 (i.e. n to the power of n). On level 4 each of the n subitems have n subitems of their own etc. (i.e. n to the power of n to the power of n to the power of…).
So we're talking about two formal grammars: FL2 which generates endlessly complex trees, and paaS which only generates a very limited tree. That's why I suggest it is best interpreted as a linear FL1 type language with the grammar S -> aaaS, where the proper parse tree can be drawn based on a morphological analysis, rather than an nonlinear FL2 language for which complex parse trees are generated by conventional formal means. How would you address this?
> [Daniel:] I think, since we're on the Conlang list, I can safely say the relevance of
these FL1/FL2 threads for most of the audience doesn't lie in the
theoretical claims, but leans more to the side of understanding the
languages proposed and what sorts of things they're good at expressing.
That's a good point. When it comes to PAS being the simplest grammar by logical necessity, I don't know what the source is. Isn't it a bit like saying there cannot possibly be any number smaller than three? Looking at the grammars aS, aaaS, abbS, abcS, aSb etc., each one looks like just another grammar. How do you know which one is supposed to be the one special one?
> For my own part, this conversation was a tangent from a conversation I'd like
to get back to about whether truly alien languages exist, or whether on the
other hand, human languages explore the set of possibilities about as
thoroughly as 7 billion people might be expected to. (So, I'm still hoping
to defend the position that truly alien languages are conceivable.) Of
course I am interested in FL2-style languages for their own,
non-theoretical sakes too.
You can rekindle the topic with a new thread, and I'll participate, although I can already admit that you're right, truly alien languages are conceivable. I just think that using sound or vision in communication seems both practical and economical, while something like using your body to create chemicals or electricity for complex communication would be far less economical. I'd expect such to be less common among the languages of the universe (when's UALS coming out?!). Perhaps it's similar to some word orders being significantly less frequent than others, in spite of being completely plausible as a means of communication :)