Language Drift: Sort-of averted for Eldraeic, due in equal parts to its origins as a designed language intended to communicate precisionist-grade thought and to its ongoing tending by the logotects, et. al., of the Conclave of Linguistics and Ontology, who are prescriptivists nonpareil. Only sort of, however, since their department of Worthy Innovations routinely combs the language as it is spoken for, well, worthy innovations to be taken up into the canonical version.
Since they’ve been doing this for a long time, and since almost nothing can ever be thrown out due to the obvious need for backwards compatibility in language design, the result tends to be — well, if it were English, they’d be like this (courtesy of xkcd):
Played rather straighter by most other languages of the Worlds, although both the influence of Eldraeic via Trade and that of the pervasive communication networks of starfaring cultures do tend to slow it down a bit.
From Academician Múírí Larathyr-ith-Lyrian, Fellow of the Sodality of Commutative Logotecture, Associate Proctor of the Conclave of Linguistics and Ontology, Loremaster of Linguistics, Semiotics, and Memetics, to the Ecumenical Commission of Translation and Conversion, greetings.
With respect to the views of the Commission and those expressed by various submissions to the commission, it remains my opinion, and that of my colleagues, that “lore” and its semiotic equivalents in other language remains the best cognate available for the Eldraeic alath. While it is in many languages of the Accord an archaic term (and thus may result in degrees of cognitive dissonance when speakers of such languages are confronted with compounds such as “spacer lore”, “nanolore”, et. al.), it is our belief that it properly reflects and thus aids in understanding the nature of the development of knowledge among the eldrae.
Unlike many civilizations whose discovery of the scientific method came as a revolutionary change of paradigm, or is perceived as such, for us the insights of Sung Iliastren and his successors formed an evolutionary phase in the search for truth; and while much of the knowledge attained by prescientific, if we may so inaccurately term them, methods was invalidated by later discoveries, we see this itself as merely part of the process of testing and refining hypotheses. Epistemology applied to itself, if you will. As such, we continue to revere the ancient scholars in fields from astronomy through chymistry to now-obsolete sorcery as fellow seekers for truth, and feel no need to discard their terminology where it remains appropriate.
I observe one of the citations offered in support of the proposal to change this translation is the various replacement terms found in the Magen dialect. While as an Imperial logotect I naturally consider this bastardized form of the language with some distaste, I would root my objection to their terminology in that the bastardized language in question belongs to a bastardized culture, which has perverted the forward-looking attitude and enthusiasm for genuine progress into a disdain for tradition and fatuous love of novelty for its own sake, hence their eagerness to replace functional words with “improvements” of no greater meaning or precision simply for the sake of doing so – something which must be rejected by any professional logotect or well-educated speaker as a matter of principle!
A third consideration is the number of related cognates (loremaster, as both a word and an academic ranking; loreworks; various trade names; etc.) which would also have to be altered in the course of execution, or otherwise lose their base root.
In closing, we must therefore reject the proposal at hand unless significant evidence of failure to understand within a sample set of educated speakers (per relevant IOSS) can be brought to our attention.
Mathematician’s Answer: The nature of the Eldraeic language, being designed as it was by mathematicians as well as logicians, philosophers, and linguists, positively encourages these. (So does the cultural attitude in favor of precision and in disfavor of the sloppy speaker and the plain intellectually inadequate.)
In fact, if asked a literally-translated “Is it A or B?” question, the only possible grammatical answers are “Yes” or “No”. If you really want to ask a question of that form – say, “Would you like tea or coffee?”, the way you say it in Eldraeic is “You would like tea how-related-to coffee?”, to which the answer is “neither…nor” (if you want nothing), “but not” (for tea), “not…but” (for coffee) or “and” (for both), any of which options can be expressed in a single conjunction.
(For a real-world example of this unusual grammatical form, this is also how Lojban does it.)
Observers of the modern Conclave on Linguistics and Ontology are unsure whether they’re saddened or delighted by how much this and other quirks of the language annoy and frustrate less precision-devoted non-native speakers, but they definitely aren’t planning on changing them any time soon…
In-universe, it’s a constructed language designed as an interlingua for the Empire by the Conclave of Linguistics and Ontology, with additional requirements for regularity, unambiguity, cultural neutrality where matters other than The Fundamentals are concerned and linguistic imperialism where they are, maximal flexibility, designed to allow the greatest scope for creativity, and simultaneously to promote logical reasoning and precision, and designed to be expressively isomorphic in multiple forms. This heady list of requirements was then tackled by a group of linguists, philosophers, logicians and mathematicians, whose work and arguments produced the language we know today.
Out of universe, it’s also a constructed language, albeit an unfinished one (and given its claims to universality, a perpetually unfinished one – I don’t have the nose to produce the olfactory-description features it inherited from the dar-bandal, for example, never mind some of the real esoterica it’s acquired from various other Starfish Languages). Nor, while I can describe in great detail its 36-character alphabet — well, 3.5 alphabets (for pen, brush, and chisel, the additional half being a variant on the brush alphabet for scratching with claws) and dozen or so phonologies (for different speech apparati, including things like radio and chromatophore matrix), do I actually have them all terribly well defined. Nonetheless, it’s constructed enough that it is possible to say things in it, and even – having participated in a conlang relay or two – for other people to understand what was said.
Type-wise, it’s somewhere between “complete original” and “foreign conversion” – originally, I started using Loglan/lojban as a base, but it’s grown up to be a very different language (it uses case tags than place structures, for one thing, along with many other affixes, and handles a lot of shared features differently, and comes with a lot of different or at least differently implemented features). And it’s nothing at all like English, certainly!