On 5/26/07, Donald J. HARLOW <[log in to unmask]> wrote:
> Hence, despite the fact that IALA as an
> organization was well aware of certain advantages of other,
> particularly schematic, planned languages (e.g. the advantage of
> having a standard adjective ending that could be applied to all nouns
> rather than having a couple of dozen different adjective endings
> whose use was etymologically determined, thus requiring, at least for
> _active_ use, memorization of adjectives derived from nouns along
> with the nouns themselves; (**)
> (**) This is, IMHO, the moral equivalent of reintroducing grammatical
> gender into a planned language, since it puts a similar burden on the
> student who wants to learn to _use_ the language rather than simply
> _recognizing_ it.

Unless, of course, the intention is to adopt this language as a
regional IAL for the Angloromance world. (This was not the agenda of
IALA, I believe, but I think it would be a good potential application
for Interlingua, given its overall design.)  Then the existence of
many adjective endings is more sensible than just one ending, as it is
very similar to the endings that exist in the native languages of the
target group of users. (At least if we consider comfort as a desirable
quality attribute. This attribute was not considered in the design of
Interlingua, but I believe the resulting language fits it well for the
target group I mentioned.)

> or the advantage, especially for
> _active_ use, of having a regular set of correlatives rather than a
> defective or even irregular set), none of these results, shown both
> by practice and by experiment (for instance, Thorndyke's famous
> experiment with the correlatives, sponsored by IALA) was taken into
> account in devising Interlingua.

Not only this was not taken into account, but also it was explicitly
ignored in the philosophy of Interlingua. Alexander Gode and Ezra
Clarke Stillman adopted a philosophy that considered only translingual
features whose presence could be objectively verified in certain
European languages, and explicitly acknowledged that they would ignore
anything else.

I believe that their fear of introducing any kind of
learnability/usability facilitator that couldn't be objectively
justified in terms of precedents in the source natlangs is a direct
consequence of the discussions promoted by IALA in the previous years.
In the first stage of the IALA history, it was observed that consensus
could not be achieved about which engineered features would be the
best for the language. Even if some features were proven to be real
learnability/usability facilitators (e.g. a regular correlative
table), they were resisted by some people because they lacked other
quality attributes perceived as desirable by some (such as apparent
familiarity and precedent in at least one natlang). So, (I believe)
Clarke thought and Gode agreed: "If subjective and arbitrary features,
even if good for a conauxlang in our subjective opinion, are strongly
resisted by some factions, then we will have to create a language
based on objective linguistical facts, this way we will make
subjective criticisms invalid. People will be able to perform valid
criticisms only when they find features in the new language that do
not comply to the linguistical facts observed in the source natlangs.
But here we will make the proviso that the language can be fixed at
any time, if a valid criticism appears." This thought guided most of
the work during the second stage of IALA's history. Of course, that
didn't stop the flux of "subjective" criticisms, but a person that
adopts the last IALA mindset will consider them as invalid, and thus
ignore them. Whether that mindset is the most adequate one is open to

In summary, the objectivity in the design of Interlingua was a
reaction to the lack of consensus in defining the priority of the
desirable quality attributes for a conauxlang. Without a well-defined
benchmark of priorities, it is not possible to design a language to
satisfy that benchmark, and it is not possible for an independent
external group (e.g. an official UNO committee) to assess whether the
language satisfies that benchmark. So, the language creators thought,
let's be extreme and consider only translingual features that can be
objectively verified, no matter if they improve or hurt learnability,
usability, regularity, cuteness, etc.

Antonielly Garcia Rodrigues