Alleged Beliefs

On my Twitter timeline, a link to an article by Katy Steinmetz at the TIME magazine’s website, titled “Elvish, Klingon and Esperanto—Why Do We Love To Invent Languages?”, came up several times this evening. In this article, Steinmetz interviews Michael Adams, professor of English at Indiana University, on conlanging. I left the following comment:

“Are invented languages better designed than natural languages?
— That’s what their inventors believe.”

Invented languages better than natural ones? Well, I suppose that’s an idealist view. They will probably never be as complex, for one, as natural languages if that’s your criterion of measuring quality. Natural languages grew and changed and diversified by means of being spoken by hundreds to billions of people over the course of millenia – a process which a single creator or even a group can never fully immitate – leaving us with a wealth of forms to explore and build our own languages on, and be it just for the love of tinkering. On the other hand, do invented languages need to be as complex as natural languages, being consciously modelled after existing languages, in order to be of good quality? Not necessarily, I think. It’s about exploring possibilities and watch how things work or play together. That’s why you build models in the first place.

And actually I’ve only now realized that the question was whether constructed languages are better in their design than natural languages, not just objectively better, as I assumed in my reply was the question. 😕 Certainly constructed languages are usually designed more consciously than natural languages, which underly an evolutionary process that’s at least partly blind (or even for the most part?). But whether design decisions by authors make constructed languages inherently better than natural ones I have doubts about. They’re the results of different processes, so it’s hard to compare.

However, just for fun and because of a couple of rather elaborate sentences and vocabulary that seemed challenging, I spent the 1½ hours after writing my comment translating the whole shebang into Ayeri, minus the quotation from the article at the beginning:

[gloss]Sa engyon narānjang vehisa ban narān suhing? Māy, neprayang adareng paranas sempayyanena. Menanyam-ikan, sa kamatong tadoy kamya narān suhing bata ang perava bananley ada-yenueri. Ang nakasyon, ang tilayon nay sa palungisayon narānye suhing naramayari menang yonangya pesan manga ling sinkyanyēa. Adareng macamley si ang ming kusangisaya ikan tadoy tianya kebay soyang-nyama yenu. Eng hapangisāra eda-macam mahaley dahasyena dilānyam nay ling sinaya sa ming vehnang narānye sitang-nana, nārya-nama cānyam veha-veha. Palunganya, ang ilta kamayon kamya narānjas suhing bananyam ban narānye vehisa sang ri vehtos miran narānyena suhing? Paronyang, adareng rapōy. Ri ming dilavāng mimānjas narānvehyaman nay ming silvvāng miranyam sirī mirāra linyayereng soyang kayvteng sitanyaley. Sā tiavāng menanyam-ikan kusangan-kusanganyeley eda-yaman.
PFOC exceed-3PN language-PL.A build-CAU good language natural? Well, suppose-1S.A that-A.INAN opinion-P idealist-GEN. one-NMLZ-DAT=very, PFOC be_as_as-3PN never complex language natural if AFOC measure-2S quality-P.INAN that=category-INST. AFOC grow-3PN, AFOC change-3PN and PFOC different-CAU-3PN language-PL natural speak-AGTZ-INST hundred billion-LOC until MOT while century-PL.LOC. That-A.INAN process-P.INAN REL AFOC can double-CAU-3SM completely never creator single or=even group. AFOC.INAN rest-CAU-3S.INAN this=process treasure-P.INAN form-PL-GEN explore-NMLZ-DAT and top REL-GEN-LOC PFOC can build-1P.A language-PL self=1P.GEN, CONC=just love-DAT build~DIM. Difference-LOC, AFOC need be_as_as-3PN complex language-PL.P natural quality-DAT good language-P build-CAU REL-A INSTFOC build-3SN.P way language-PL-GEN natural? Believe-1S.A, that-A.INAN necessary-NEG. INSTFOC can explore-2S.A possibility-PL.P language-build-PTCP-NMLZ and can see-2S.A way-DAT REL-DAT-INST function-3S.INAN thing-PL-A.INAN or together-3P.INAN each_other-P.INAN. CAUFOC create-2S.A one-NMLZ-DAT=very double~DIM-PL-P.INAN this=reason.[/gloss]


  • Causative marking on verbs and the resulting meaning is still nicely irregular: Sa palungisayon (PFOC different-CAU-3PN) is supposed to mean ‘they are differentiated’, while ang ming kusangisaya (AFOC can double-CAU-3SM) is supposed to be ‘they can copy/immitate’.
  • Reduplication is fun.
  • I think I’m going to allow concessive adverbials in sentence-initial position, like English does.
  • Numbers still are a bit odd for me to work with: menang yonangya pesan (12² 12⁸-LOC until) ‘hundred to billion’ as an attributive phrase, with yonangya, although not nominalized, marked for the locative case demanded by the postposition. If I did nominalize it, the resulting meaning would be ‘billionth’.
  • The question pronoun for ‘how, in which way’ (simin) should not be used as a relative pronoun, at least not in more formal language. Instead, use miran sirī (way REL-Ø-INST) ‘the way in which’, which is also how I arrived at simin.
  • Adams, Michael. Interview by Katy Steinmetz. “Elvish, Klingon and Esperanto—Why Do We Love To Invent Languages?” TIME. 2011. Time, Inc., 16 Nov. 2011. Web. 16 Nov. 2011.