Justin B Rye 23-Jul-02
POLITICS MAIN ROOT-CLASSES

Ranto Appendices - T

GROUNDPLANS

Faced with the charge that the design of Esperanto's grammar is parochial, the nearest thing Zamenhof's apologists have to counter-evidence is the fact that Esperanto's morphology was avowedly influenced by "agglutinative" languages such as Turkish rather than the "fusional" model dominant in Europe.  What this means is that where Italian verbs have endings such as <-ai>, which signals past-tense-first-person-singular in one indivisible blob, the Turkish equivalent is <-d-im>, where the <-d-> marks the tense and the <-im> carries the person agreement.  It was recognised well before Zamenhof came along that this makes a better groundplan for auxlang morphology, since it avoids the need to memorise combinatorial tables of grammatical endings.  In Zamenhof's neighbourhood this groundplan was represented by Finnish, Estonian and Hungarian - all predominantly agglutinative, though they look very westernised when put alongside other examples such as Korean or Quechua.

Zamenhof eagerly adopted the concept of discrete invariable building-blocks; but there are two kinds of block - "derivational", used to build new vocabulary items, and "inflectional", used to signpost syntactic features.  It's mostly the derivational affixes that Esperanto links together in long but regular chains, as in the compound noun <te-krucho-mufo-kolekt-ist-ar-ejo> "tea-pot-cosy collectors' club-house".  The inflectional ones never get much beyond <ekzil-it-o-j-n> "exiles" - and even there the <-ojn> is designed to merge into a single unit.  In effect Esperanto is like a version of German with its affixes de-fused, not like a paradigmatically agglutinative language.  A whole-heartedly Turkish-style auxlang would handle all the various modal, reflexive, conditional or aspectual forms of verbs by stacking verb-endings, so that (for instance) "I won't have been seen", <mi ne estos vidita>, would instead use perfective, passive, future and negative suffixes to form something like, say, <mi vid-iv-at-ur-en>.

The third major option (which has influenced Esperanto's verbal system there) is the "isolating" groundplan, which consists of eliminating affixes ("re-straight-en-ed") in favour of multiword phrases ("did make straight again").  It turns out that a case can easily be made for thoroughly isolating solutions being more convenient for more people:

• They have better extensibility - that is, they can be taken from sparse simplicity to rich subtlety just by learning more vocabulary.
• They accommodate different agreement-marking instincts well - you can include or omit or relocate (e.g.) aspect-marking words depending on your mother-tongue habits (and there's nothing to stop you using "did-make-straight-again" as if it was a single agglutinated word).
• Affixing grammars are basically alien to speakers of languages with no affixing features; but the barrier is one-way, because there are no languages entirely lacking in isolating features.
• Isolating languages are by no means obscure - they dominate East Asia from China to Vietnam to Indonesia, and English is largely isolating itself; that adds up to almost a third of the human race already!

The natural equivalent of the constructed auxiliary language is the "creole" (which is what a sub-linguistic pidgin turns into once children start growing up as native-speakers); they are "designed" by the innate preference babies have for a complete but easily learnable grammar, and they tend overwhelmingly to use isolating rather than agglutinating groundplans.

(And for the sake of completeness I should also mention the fourth basic groundplan: "polysynthetic" grammars are exemplified by the West Greenlandic one-word sentence <ininnukalaarniarlungaana>, "the thing is, I'm going to my room for a bit"... but this is rarely proposed as a model for an auxiliary language!)