[Disclaimer: I am no good at all in Sostematiko and fairly good in toki pona. This surely colors my view of things when I step away from "objective facts."]
One of the virtues of toki pona (tp) is that it has brought the search for minimal languages into focus: any such search must now make explicit comparisons with tp, and most acknowledge tp as an inspiration. tp is not a perfect minimal language, of course, It famously could make do with two fewer vowels with a single vocabulary change and with two fewer consonants using only a couple minor changes in phonological rules.Its vocabulary of 120 (give or take) words contains at least one redundancy and a couple of cases with dubious claims to be in a minimal set. Even some cases in its already restricted grammar could easily be handled within the slightly modified general rules. And, of course, there are some doubts about how much can be said in tp -- enough to claims "universal" coverage as the idea of a minimal language requires. But tp is still the standard.
Sostematiko arose from the quest by the author, Giannes Kenanides (he writes it as Giannhs Kenanidhs to stress the long e sounds) for a simplified Greek, now presented as Free Greek. Starting from the limited basic vocabulary and grammar of that project, he moved to a language separated from the particularities of Greek (aside from vocabulary) to find a general minimal language which would overcome the perceived shortcomings of tp.
The phonology of Sostematiko is clearly non-minimal in absolute terms since it adds to that of tp the voiced versions of the stops and s: /b, d, g/ and /z/, and also /f/ and /x/ (voiceless velar fricative, Grm ch), tp /l/ is replaced by /r/. In all, five vowels (as in tp) and thirteen consonants (against 7). Sostmatiko syllables are much more complicated than those of tp, since both consonant clusters, initial and final, and diphthongs are permitted, Stress is not specified nor marked, Word length varies indefinitely, and, indeed, the official notion of a words is not fixed, since the language is claimed to be interpretable as isolating, agglutinating or paradigmatic (but the actual presentatrion differs somewhat in these different forms).
The vocabulary is entirely Greek, though a mixture of various stages and subject to processes of simplification (the distinction between long and short vowels is lost, for example, and that between theta and tau and lambda and rho). There are 220 root words (against tp's 120) but, as in Esperanto, these rarely occur alone in a text, but rather come with additional bits marking nouns, active and passive verbs and adverbs, as well as other affixes, both grammatical (subjects, plural, object) and derivational (though all in theory capable of occurring, in somewhat different guises, as separate words). One can take this as adding 100 words to tp or as multiplying tp;s total by up to a factor of 10, On the other hand, they are, for the most part, somewhat familiar, turning up in English words with familiar meanings (even if not always obvious).
In terms of minimality and general language efficiency, the former view is probably fairer. While the number of separate words (in the usual format) is much greater, the grammar is simplified but keeps a close parallel with tp. Thus, Sostematiko is basically SVO, like tp, but does not need either 'li' for V or 'e'' for O, since (in the basic case) everything before the first verb (marked by '-i' or '-eti' is Subject) and the noun after it (marked, like the subject, with the '-o' of noun) is Object, Similarly, in at least some dialects, the difference between adjectival and nominal modifiers of nouns is marked, eliminating an ambiguity in tp, where, for example, 'ilo jan' might be either a human machine (robot) or a person's tool, depending upon whether 'jan' is taken as (as we would say in a language with nouns and adjectives as normative) a noun or an adjective. Further, the existence of passives -- for all that allows the avoidance of responsibility ploy -- avoids the frequent roundabout constructions full of dummy references that are needed in tp.
While Sostematiko is SVO and AN, it allows for SOV forms (using the optional '-n' object). In particular, it allows relative clauses among the noun modifiers and places them in front of their noun, typically using the SOV - perhaps even OSV -- form to mark the boundaries, relative and verb. Having relative clauses rather than separate sentences complicates the grammar but arguably improves rapid understanding (especially since, in tp, the difference between restrictive and nonrestrictive relative clauses is not well represented). The variant orders also allow disambiguating cases where verb phrases directly modify nouns (OVS as a noun phrase), which are ambiguous in tp.
Sostematiko allows roots to be compounded, eliminating another ambiguity of tp. In tp, word strings may well come to have a special meaning as a whole: 'jan pona' as "friend" , to take the stock example. But they always remain word strings and so capable of still having their basic meaning, "good person', in this case. Fusing the idioms into words (the mythical 'janpona', /jambona/) would eliminate an array of alternate readings with low probabilities and never intended (preediting does that work now, but is, of course, unreliable).
Yet these advantages come at a cost in learning and reading. Once one has learned the vocabulary of tp, one never has to puzzle out the words of the text, whereas Sostematiko words often need analysis before one can go on to place them correctly in the sentence. Even if the analysis is simple, the fact is that the words often cannot be recognized at a glance (or even in typoed format) as they can in tp. There even appear to be some words that have false, but plausible, derivations from reasonable but not actual words, which all leads to chases through the dictionary and the creation of new words.
Many of the problems I have experienced with Sostematiko (and I have to say that, while it is apparently ahead of tp on points as useful language, I find it nearly impossible to use) I attribute to the present lack of even moderate teaching material, say anything as good as the various lessons for tp. Learning it is like trying to get Esperanto with only a glossary and the 16 rules but no -- or very few -- bits of expounded text or examples. There are a handful of texts with translation aids, but little in the way of general principles of interpretation. As a result, each new phrase that is not totally literal is a puzzle to be solved. I've found that, where possible, translating into tp helps quite a bit ('ano membrano' , "up sheet" becomes 'lipu sewi' and immediately recognizable as "holy book". which at least fits the context, though may not turn out to be right). But such translations are not always possible and are subject still to the different spirits of the languages. The problem with Sostematiko is that so little of its spirit is yet revealed. As it emerges, perhaps Sostematiko will become a major contender in the small-language field.
One of the virtues of toki pona (tp) is that it has brought the search for minimal languages into focus: any such search must now make explicit comparisons with tp, and most acknowledge tp as an inspiration. tp is not a perfect minimal language, of course, It famously could make do with two fewer vowels with a single vocabulary change and with two fewer consonants using only a couple minor changes in phonological rules.Its vocabulary of 120 (give or take) words contains at least one redundancy and a couple of cases with dubious claims to be in a minimal set. Even some cases in its already restricted grammar could easily be handled within the slightly modified general rules. And, of course, there are some doubts about how much can be said in tp -- enough to claims "universal" coverage as the idea of a minimal language requires. But tp is still the standard.
Sostematiko arose from the quest by the author, Giannes Kenanides (he writes it as Giannhs Kenanidhs to stress the long e sounds) for a simplified Greek, now presented as Free Greek. Starting from the limited basic vocabulary and grammar of that project, he moved to a language separated from the particularities of Greek (aside from vocabulary) to find a general minimal language which would overcome the perceived shortcomings of tp.
The phonology of Sostematiko is clearly non-minimal in absolute terms since it adds to that of tp the voiced versions of the stops and s: /b, d, g/ and /z/, and also /f/ and /x/ (voiceless velar fricative, Grm ch), tp /l/ is replaced by /r/. In all, five vowels (as in tp) and thirteen consonants (against 7). Sostmatiko syllables are much more complicated than those of tp, since both consonant clusters, initial and final, and diphthongs are permitted, Stress is not specified nor marked, Word length varies indefinitely, and, indeed, the official notion of a words is not fixed, since the language is claimed to be interpretable as isolating, agglutinating or paradigmatic (but the actual presentatrion differs somewhat in these different forms).
The vocabulary is entirely Greek, though a mixture of various stages and subject to processes of simplification (the distinction between long and short vowels is lost, for example, and that between theta and tau and lambda and rho). There are 220 root words (against tp's 120) but, as in Esperanto, these rarely occur alone in a text, but rather come with additional bits marking nouns, active and passive verbs and adverbs, as well as other affixes, both grammatical (subjects, plural, object) and derivational (though all in theory capable of occurring, in somewhat different guises, as separate words). One can take this as adding 100 words to tp or as multiplying tp;s total by up to a factor of 10, On the other hand, they are, for the most part, somewhat familiar, turning up in English words with familiar meanings (even if not always obvious).
In terms of minimality and general language efficiency, the former view is probably fairer. While the number of separate words (in the usual format) is much greater, the grammar is simplified but keeps a close parallel with tp. Thus, Sostematiko is basically SVO, like tp, but does not need either 'li' for V or 'e'' for O, since (in the basic case) everything before the first verb (marked by '-i' or '-eti' is Subject) and the noun after it (marked, like the subject, with the '-o' of noun) is Object, Similarly, in at least some dialects, the difference between adjectival and nominal modifiers of nouns is marked, eliminating an ambiguity in tp, where, for example, 'ilo jan' might be either a human machine (robot) or a person's tool, depending upon whether 'jan' is taken as (as we would say in a language with nouns and adjectives as normative) a noun or an adjective. Further, the existence of passives -- for all that allows the avoidance of responsibility ploy -- avoids the frequent roundabout constructions full of dummy references that are needed in tp.
While Sostematiko is SVO and AN, it allows for SOV forms (using the optional '-n' object). In particular, it allows relative clauses among the noun modifiers and places them in front of their noun, typically using the SOV - perhaps even OSV -- form to mark the boundaries, relative and verb. Having relative clauses rather than separate sentences complicates the grammar but arguably improves rapid understanding (especially since, in tp, the difference between restrictive and nonrestrictive relative clauses is not well represented). The variant orders also allow disambiguating cases where verb phrases directly modify nouns (OVS as a noun phrase), which are ambiguous in tp.
Sostematiko allows roots to be compounded, eliminating another ambiguity of tp. In tp, word strings may well come to have a special meaning as a whole: 'jan pona' as "friend" , to take the stock example. But they always remain word strings and so capable of still having their basic meaning, "good person', in this case. Fusing the idioms into words (the mythical 'janpona', /jambona/) would eliminate an array of alternate readings with low probabilities and never intended (preediting does that work now, but is, of course, unreliable).
Yet these advantages come at a cost in learning and reading. Once one has learned the vocabulary of tp, one never has to puzzle out the words of the text, whereas Sostematiko words often need analysis before one can go on to place them correctly in the sentence. Even if the analysis is simple, the fact is that the words often cannot be recognized at a glance (or even in typoed format) as they can in tp. There even appear to be some words that have false, but plausible, derivations from reasonable but not actual words, which all leads to chases through the dictionary and the creation of new words.
Many of the problems I have experienced with Sostematiko (and I have to say that, while it is apparently ahead of tp on points as useful language, I find it nearly impossible to use) I attribute to the present lack of even moderate teaching material, say anything as good as the various lessons for tp. Learning it is like trying to get Esperanto with only a glossary and the 16 rules but no -- or very few -- bits of expounded text or examples. There are a handful of texts with translation aids, but little in the way of general principles of interpretation. As a result, each new phrase that is not totally literal is a puzzle to be solved. I've found that, where possible, translating into tp helps quite a bit ('ano membrano' , "up sheet" becomes 'lipu sewi' and immediately recognizable as "holy book". which at least fits the context, though may not turn out to be right). But such translations are not always possible and are subject still to the different spirits of the languages. The problem with Sostematiko is that so little of its spirit is yet revealed. As it emerges, perhaps Sostematiko will become a major contender in the small-language field.
No comments:
Post a Comment