On Tue, Jan 8, 2013 at 9:08 PM, And Rosta <[log in to unmask]> wrote:

>> Now, if you are thinking "learnable by humans or similar beings", then I
>> would agree -- it is possible to have a language that human beings would
>> not be able to learn.  But I don't think that's the point of that design
>> feature -- it's not that it's "learnable" by any specific entity, it's
>> that
>> it is transmitted through cultural learning rather than genetically.
> I think you characterize that criterion correctly. But the question is: In
> a universe in which English was spoken only by robots in whom the rules of
> English were hardwired, would English not be a language? In my view,
> English is English, regardless of how its speakers come to know it.

 Ah, but if the whole language is hard-wired, then the robots would be
unable to coin new words, which hampers productiveness.  I think the whole
issue of language being culturally learned dovetails with the productivity
feature.  As far as whether your hypothetical robots are speaking English
-- it's tricky.  Do current voice recognition systems like Siri
speak/understand English, albeit imperfectly?  I don't know much about the
programming of those systems, but my guess is that most of what they do is
find key words and do specific functions based on those.  But even with the
yet-to-be developed technology for computers to truly parse and comprehend
human grammar, I think the ability to learn new words and structures will
be necessary for these systems to really have language as humans have it.

It's an interesting gray area to explore, and we are only going to see more
and more real-world scenarios approaching this particular philosophical

On Tue, Jan 8, 2013 at 10:11 PM, Patrick Dunn <[log in to unmask]> wrote:

> But Hockett's principles (which is what we're working on tomorrow, in an
> abbreviated version -- hey, there's finally an 11th edition of Language
> Files!) don't strike me as all ideal (in the Platonic sense).  For example,
> a perfectly good system of communication might exist that isn't arbitrary;
> but if it does, it's not human.  Similarly, very effective systems of
> communication exist that aren't productive.  Bees, for example.  But if
> linguists had to spend their time looking at bees wiggle, we'd be
> entomologists (and I'm more interested in etymology).  So by saying
> "productive!" we can get rid of a lot of communication systems that aren't
> human, and focus on one that is.

Is the goal to find the communication system humans use, or study this
particularly complex communication system that we just happen to only
observe in humans.  I really feel there's something more to language as a
system than just "humans do it and nothing else".  That is a quite
fascinating part of the issue, and it is interesting that no other animal
has achieved this incredibly evolutionary advantage (to our knowledge), but
I feel like just saying "what do humans do, how do we exclude everything
else" isn't totally the right way to define things.  We might yet find some
loquent aliens somewhere.