I agree with you, as can be seen by comparing my original message to Miles Forster about LIFO conlangs.

In haste,

--And.

Logan Kearsley, On 19/04/2012 00:53:

> On 18 April 2012 17:19, And Rosta<[log in to unmask]>  wrote:
> [...]
>> Logan writes: "if a general LIFO stack is not only sufficient but necessary,
>> then the human brain can't handle it, because language doesn't count higher
>> than 2. Natural language processing abilities are neither a strict superset nor subset of
>> what an ideal push-down automaton can handle; we're differently constrained
>> by having small, finite memory."
>>
>> The stack as I have described it does not require counting, and its
>> vulnerability to the limitations of short-term memory is in fact key
>> evidence for its involvement in psycholinguistic parsing.
>
> It only doesn't require counting if you never have to explicitly
> access something at a depth other than "on top". And if that's the
> case, the fact that you happen to be using a stack underneath is not
> particularly interesting, because you could just as well use something
> else and get the same results.
>
>> Logan continues:
>>
>> "When discussing the idea of LIFO grammar in a conlang context, it's
>> important to distinguish between a language that *can* be parsed by a
>> stack machine, and a language that *relies* on being interpreted by a
>> stack machine. The former I consider to be a rather useless
>> categorization, but Fith is a LIFO language in the second, stronger
>> sense. It may thus be less computationally complex than human
>> languages, but falls prey to the division between "problems that our
>> brains can be trained to solve by conscious emulation" and "problems
>> for which our brains have a suitable algorithm compiled in"."
>>
>> Again to repeat the earlier thread, there is no essential difference between
>> Fith (without stack operators) and English. The tree structure for a Fith
>> sentence would be indistinguishable from the tree structure of a head-final
>> natlang.
>
> Fith without stack operators is not Fith.
> The result of removing the stack operators from Fith isn't anything
> that I would find it useful to classify as a LIFO stack language,
> precisely because it is indistinguishable from a normal head-final
> natlang; it doesn't *do* anything with the fact that it happens to be
> implemented with a stack.
>
> -l.