In defence of philosophical languages (was: RE: Comparison of philosophical languages

Sally summarizes the usual & valid objections to philosophical
languages:
> It was a tragic failure.  The seventeeth-century language philosophers
> erred in thinking that they could replace the arbitrary and therefore
> flawed system of natural language by a language that perfectly reflected
> ratiocination and taxonomy (hence "philosophical").  Their words were
> meant to DESCRIBE the nature of the things they symbolized and their
> place within a huge map of listed items.  First, Wilkins wrote an enormous
> system into which he tried to put "universal" thought (not understanding
> that his thought was basically inflected by seventeenth-century English
> language and society).  Then he devised characters ("real characters") that
> would symbolize the basic elements of those thoughts, uttered in single
> syllables, and which could be arrived at if you understood his complicated
> taxonomy.  Then he applied what he supposed were the most rational sounds
> to represent those syllables of thought.  His error was that that is
> exactly not how language works.  All his root vegetables were expressed
> in words that sounded too much alike, once you climbed down the ladder of
> his system Real language operates in our minds through difference, and
> its symbols are ultimately arbitrary in origin: radish, carrot, turnip,
> beet, parsnip, rutabega... these words have many different origins, and
> they are ultimately easier to distinguish than elevela, elevale, elevali,
> elevalo, elevalily, etc.  Wilkins' system also allowed for no neologisms,
> and ultimately his language was just too difficult to learn.
Yes, a language needs room for neologisms. Yes, a language that embodies
a purely taxonomic map of meaning is unsatisfactory. But...

Any language is partly a map of conceptual space, a kind of index to
an encyclopedia of thought. It seems a virtue for the language to
aspire to make its mapping visible to the inspection of the ordinary
speaker and not just to the scholar.

A language (therefore) reflects a culture, and Wilkins's reflects the
rationalistic culture of his time.

Had Wilkins's classificatory scheme been informed by modern knowledge
of semantic typology, it could have had more legitimate claim to be
quasi-universal -- not by the standards of universal scientific
'truth' but by the standards of cognitive universals.

Differences between things, such as carrots and potatoes, need to
be reflected in wordshape by phonological differences that reflect
the unconfusability of carrots and potatoes. But at the same time,
there are reasons for wanting to have words with similar or related
meaning have similar or related sound. One reason is that an iconic
map of conceptual space is a better (more faithful) map. Another
reason is that the iconicity makes the language easier to learn.
Another reason is that there is a universal tendency in language to
develop iconicity: we see it in compounding, in phonaesthemes, and
in sets like {north, south}, {east, west}, {female, male}, {dog,
hog, frog, polliwog (tadpole)}.

I conclude that Wilkins's effort was not bad relative to the
state of knowledge of his time, and his efforts, rather than
discrediting his entire enterprise, allow us to learn from his
mistakes yet still attempt a project that preserves his goals
to some extent.

So a modern philosophical language would have the following
properties:
* Its model of basic/major categories would be based on the study
of natural language semantics.
* Words with similar/related meanings should in shape be half
similar and half different. But the different half may be similar
to other words related in a different way. For example, the
shape of the word for carrot might have a tubery component in
common with the word for potato, and an orangey component in
common with copper, red hair, fire, etc., much as, say 'crunch'
has elements of meaning in common with other cr- words and
other -unch words.
* The more systematically analysable a semantic field is (within
the constraints of cognitive naturalism), the more systematic
the sound--meaning iconicities can be.

--And.

Replies