[YG Conlang Archives] > [romconlang group] > messages [Date Index] [Thread Index] >
Hi! Jan van Steenbergen writes: >... > > - I make a terminological distinction between 'versions' in > > real time and 'stages' in imaginary time meant to provide > > orientation when exploring the development through real > > time of the imaginary history of imaginary languages, > > where one has to deal with two dimensions of time: > > That difference is quite essential, yes. Note that conlangers go > about that different. Someone, IIRC Nik Taylor in his Uatakassi, uses > earlier versions of the language as dialect, and I vaguely remember > that someone else uses different versions of the language as > different stages in its con-history. That would be me for some languages, e.g. Fukhian. For others, e.g. Þrjótrunn and Qþyn|gài, I follow your approach: > That is something I have never done myself. Older versions of Wenedyk > or any other of my languages are just older versions, now incorrect > and no longer valid. For me, it mostly has to do with the way of writing and revising the documentation: most newer conlangs with a reasonable grammatical complexity have a Lisp-written grammar and all texts in the documentation source code are represented as an abstract declaration which is converted into the conlang by the Lisp grammar (which also does the lexicon lookups via SQL). This makes revisions easy: all texts change immediately without the need to edit old texts. For such conlangs, the only old texts appear outside the scope of the main document, say in Relays. These are then tagged as old language versions. However in Fukhian, for example, the documentation itself contains unrevised, old examples, because I found it hard at that time to keep in sync with the current grammar definition. Because I don't like outdated documentation, I simply defined the sentences to be older in conhistory. You could call it cheating. :-) OTOH, some interesting conlang changes are still documented and did not silently disappear, e.g. the drop of one of the case endings when postpositions are added. **Henrik