Wiktionary talk:About Algonquian languages

comment which had previously been on the main page
I suppose this section would make the most sense to address my question. What about the issue of a template for Algonquian languages, such as Ojibwe, where we can either list every single word ever printed since the time of European Contact and still not cover all the words or instead listing the existing "preverb", "postverb", "prenoun" and "postnoun", along with true modifiers. These "pre..." and "post..." by themselves are not necessarily words, but they don't fall under "etmology" since they each do have their own etmologies. In addition, these parts may experience certain grammatical transformation, have different classes of gender- and proximity-affected conjugations, etc., etc. Think of it like a language made up of nothing but prefixes and suffixes where each combination also can act as a prefix or suffix of another word to create yet another word. Oh, and there are actual prefixes and suffixes that are not "preverb", "postverb", "prenoun" or "postnoun".

==Ojibwe== ===Etymology=== ===Transformation=== ===Pronunciation=== ===#form 1=== ====Adjective=== =====Derived Adjective===== =====Derived Noun===== =====Derived Verb===== =====Derived Adjective===== =====Derived Noun===== =====Derived Verb===== ====Noun==== =====Derived Adjective===== =====Derived Noun===== =====Derived Verb===== ====Verb==== =====Derived Adjective===== =====Derived Noun===== =====Derived Verb===== ===#form 2=== ====Adjective=== ====Noun==== ====Verb==== ===#form 3=== ====Adjective=== ====Noun==== ====Verb==== ===#form 4=== ====Adjective=== ====Noun==== ====Verb====
 * 1) form 1
 * 2) form 2
 * 3) form 3
 * 4) form 4
 * 1) etym 1
 * 2) etym 2
 * 1) trans 1
 * 2) trans 2
 * 1) pron 1
 * 2) pron 2
 * 1) meaning 1.1 (e1, p2)
 * 1) meaning 1.1a.1 (e1, p2)
 * 2) meaning 1.1a.2 (e2, p2)
 * 1) meaning 1.1a.1 (e1, p2)
 * 1) meaning 1.1a.1 (e2, p1)
 * 2) meaning 1.2 (e2, p2)
 * 1) meaning 1.2a.1 (e1, p2)
 * 2) meaning 1.2a.2 (e2, p2)
 * 1) meaning 1.2a.1 (e1, p2)
 * 1) meaning 1.2a.1 (e2, p1)
 * 1) meaning 1.1 (e1, p1)
 * 1) meaning 1.1n.1 (e1, p2)
 * 2) meaning 1.1n.2 (e2, p2)
 * 1) meaning 1.1n.1 (e1, p2)
 * 1) meaning 1.1n.1 (e2, p1)
 * 1) meaning 1.1 (e2, p1)
 * 1) meaning 1.1v.1 (e1, p2)
 * 2) meaning 1.1v.2 (e2, p2)
 * 1) meaning 1.1v.1 (e1, p2)
 * 1) meaning 1.1v.1 (e2, p1)
 * 1) meaning 2.1 (e1, p2)
 * 2) meaning 2.2 (e2, p2)
 * 1) meaning 2.1 (e1, p1)
 * 1) meaning 2.1 (e2, p1)
 * 1) meaning 3.1 (e1, p2)
 * 2) meaning 3.2 (e2, p2)
 * 1) meaning 3.1 (e1, p1)
 * 1) meaning 3.1 (e2, p1)
 * 1) meaning 4.1 (e1, p2)
 * 2) meaning 4.2 (e2, p2)
 * 1) meaning 4.1 (e1, p1)
 * 1) meaning 4.1 (e2, p1)

As you can see, using one of the template example several postings above to fit into the Ojibwe structure, it has already branched into something very frightening. If Ojibwe were to fit the Indo-European types of template, Ojibwe has the potential to just simply clog up computer resources. If an all-encompassing template were to be created, there should be a way to bring all this under control or specific guide-lines ought to be given so that we don't go on and on ad nauseum... like what potentially can happen with Ojibwe or any other Algonquian language, or Algonquian languages with their near-infinite word permutation capabilities don't take up all the server resouces for listing out every single possible word (and it associated conjugation and derivatives). I began updating gichi-, maajii- and moose nearly 2 months ago but have stopped until a more-agreeable template is made available such that the possible infinite permutation of word list would not happen. Feed back on how to proceed would really would be appreciated.CJLippert 23:36, 3 January 2006 (UTC)

Maybe what would work better is if we don't even divide the words into grammatical classes and just list them by their form classes. If we have the first level of template as this:

==Algonquian Language Word== ===Etymology=== ===Transformation=== ===Pronunciation=== ===#form 1=== ===#form 2=== ===#form 3=== ===#form 4===
 * 1) form 1
 * 2) form 2
 * 3) form 3
 * 4) form 4
 * 1) etym 1
 * 2) etym 2
 * 1) trans 1
 * 2) trans 2
 * 1) pron 1
 * 2) pron 2
 * 1) word 1.1
 * 2) word 1.1.1
 * 3) word 1.1.2
 * 4) word 1.1.3
 * 5) word 1.2
 * 6) word 1.3
 * 1) word 2.1
 * 2) word 2.2
 * 3) word 2.3
 * 1) word 3.1
 * 2) word 3.2
 * 3) word 3.3
 * 1) word 4.1
 * 2) word 4.2
 * 3) word 4.3

and have the second level of template as this:


 * 1) word 1.1
 * 2) grammatical class 1 {link to gramatical class conjugation/declension table}
 * 3) definition 1
 * 4) definition 2
 * 5) definition 3
 * 6) grammatical class 2 {link to gramatical class conjugation/declension table}
 * 7) definition 1
 * 8) definition 2
 * 9) grammatical class 3 {link to gramatical class conjugation/declension table}
 * 10) definition 1
 * 11) definition 2

The grammatical classes probably ought to be listed by verbs, nouns, pronouns, particles; nested in those should be gender ordered as inanimate then animate; nested yet within that transitivity by intransitives then transitives; further nested with directionality.

So, for example, in Ojibwe if we use zhiishiigwe, the results would be something like this:

= Ojibwe = == ᔒᐦᔒᐝ (zhiishii) == * < PA: šîhšîy (gourd/rattle) * "rattling" indicator CJLippert 00:53, 29 January 2006 (UTC)
 * 1) ᔒᐦᔒᐝᑯ (zhiishiigo)
 * 2) * < ᔒᐦᔒᐝᒄ (zhiishiigw=)
 * 3) * {vii}
 * 4) ** ST rattles
 * 5) ᔒᐦᔒᐝᑴ (zhiishiigwe)
 * 6) * {vai}
 * 7) ** SB rattles
 * 8) * {na}
 * 9) ** rattlesnake
 * 10) ᔒᐦᔒᐝᑴᒌᐱᐦᒃ (zhiishiigwejiibik)
 * 11) * < zhiishiigwe (rattlesnake) + jiibik (root)
 * 12) * {ni}
 * 13) ** lion's foot (plant of genus Prenanthes)
 * 14) ᔒᐦᔒᐝᑴᐗᐦᔮᓐ (zhiishiigwewayaan)
 * 15) * < zhiishiigwe (rattlesnake) + wayaan (skin)
 * 16) * {ni}
 * 17) ** rattlesnake skin (not on a rattlesnake)
 * 18) ** rattlesnake skin bag
 * 19) * {na}
 * 20) ** rattlesnake skin (on a rattlesnake)
 * 21) ᔒᐦᔒᐝᑾᓐ (zhiishiigwan)
 * 22) * {vti}
 * 23) ** SB rattles ST
 * 24) * {ni}
 * 25) ** rattle
 * 26) ᔒᐦᔒᐝᑾᓂᕽ(Zhiishiigwaning)
 * 27) * < zhiishiigwan (rattle) + {loc}
 * 28) * {ni}
 * 29) ** Sheshegwaning (Reserve #20), Ontario
 * 30) ᔒᐦᔒᐝᑾᓂᕽ ᐊᓂᐦᔑᓈᐯ (Zhiishiigwaning Anishinaabe)
 * 31) * < zhiishiigwaning (Sheshegwaning) + anishinaabe (Indian)
 * 32) * {na}
 * 33) ** Sheshegwaning First Nation (ON)
 * 34) ᔒᐦᔒᐝᑾᓇᓐ (zhiishiigwanan)
 * 35) * {vta}
 * 36) ** SB rattles SB

Category:Algonquian languages Algonquian

About Algonquian languages
At this stage we really don't have a lot of people who are competent in any Algonquian language, so I'm sure you are in an excellent position to influence how such languages will be treated in the future. I absolutely agree that the traditional Indo-European concepts of "noun" and "verb" may not be fully transportable outside of that family of languages. Your approach could still be changed at some time in the future, but that could be difficult when some sufficiently knowledgeable person comes along. I gather from your comments that this format is suggested for all Algonquian languages, and not just Ojibwe. This is why I did not move the comments to Wiktionary:About Ojibwe. I notice that you prefer the spelling "Ojibwe", whereas Ethnologue prefers "Ojibwa", and the form "Ojibway" seems to have passed off into history. Can you provide some background to this decision? Ethnologue shows Ojibwe as really being 8 different languages. I also see that you are in Minnesota where the Chippewa language is the first nation language that appears to prevail. How different are these 8 languages, and what allowances should we be making for these differences? Eclecticology 22:14, 9 January 2006 (UTC)
 * That all depends on how inclusive or exclusive one is. If a language is defined as language used by a recognized government, there would be over 200 languages in the Ojibwe group because each community is a Nation unto itself.  If one defines the language as the Ojibwe-group itself of the Algonquian languages, then there is only one.  At least in Freelang Ojibwe Dictionary, as a reflection of Ojibwe Language Society-Miinawaa, the broadest grouping was chosen and then by geographical, political and traditional alliance reasons divided the language up into 10 dialects comprised of 19 sub-dialects (the sub-dialects were not specifically pointed out), not 8 languages like Ethnologue.  Though in English "Ojibwa" is the official spelling, in the language of the Ojibwa, use the Fiero romanization has been specifically sanctioned as a way for communicating internationally, and the spelling of the word is "Ojibwe"... sort of like "German" vs. "Deutsch"  The spelling of "Ojibway" is common because otherwise those who are not familiar with the "Ojibwa" spelling's pronunciation would incorrectly say "o-jib-wuh" and those who are not familiar with Fiero would say "o-jib-wee".  In the broadest sense, the languages of the Ojibwe-group are Anishinaabemowin (Algonkin, Mississauga, Ojibwa, Nipissing, Saulteaux, Odawa and Potawatomi) and Anishininimowin (Oji-cree), with Oji-cree being the transition language/dialect between Anishinaabemowin and Cree, while Algonkin being the transition language between Anishinaabemowin and Abenaki.  Since the Algonkin call themselves "Anishinaabe", Freelang included them.  Though Oji-cree calls themselves "Anishinini", because they are mutually intelligible, they have been included as well.  Ethnologue completely separates Potawatomi out, while in Freelang Ojibwe, only the Southern Potawatomi has been excluded, with Northern Potawatomi being the transition between the rest of Ojibwe-group from the Southern Potawatomi.  All Algonquian languages are polysynthetic and are agglutinating.  As shown in the example, one can define every single permutation of the word and still not cover all, or one can define the parts and let the user assemble the words.  However, some combinations are not as obvious, so more common assemblences must be provided.  Freelang Ojibwe is modelled after Concise Dictionary of Minnesota Ojibwe because of two simple reasons: 1) it was originally a digital version of that book but it grew as more dialects were added in, and 2) agglutinated words are given because when Ojibwe-English dictionaries and wordlist were made, that is what was expected during the early Christian Missionary period, and often time, still is.  Notable exeption to this is the Lexique Algonquin by Cuoq, who by breaking the words down to their basic pieces, showing the permutations of those basic pieces, basically allowed the users to assemble the words for themselves.  A Wiktionary should strike a good balance between those two presentation philosopies.  However, for the case of Anishinaabemowin since that time, the syncopated accenting system in some dialects have become so pronounced that vowels, and sometimes whole syllables, were dropped if unaccented (such as with Odawa, Potawatomi and Mississauga).  Other Algonquian languages have this syncopated accenting system, but they don't experience out-right syncope.  Features common to all Algonquian languages are the need for animate/inanimate gender and proximities, for both nouns and verbs.  Others are verb transitivity, topic directionals, number, voice and modes.  True nouns also need actor persons, number, dependencies, modes.  True adverbs require modes and number.  True pronouns have dependencies, proximities, number and gender.  Within each Algonquian language, a word can be decomposed into their root part.  Each of those root parts have their etmology.  Consequently, I can say "zhiishiigwe" and "zhiishiigwan" in Ojibwe have the same root, an unglossed verb-root "zhiishiigw=" to mean "to rattle" resulting in:
 * zhiishiig (vii) -- ST rattles ==> this too is unglossed
 * zhiishiigwe (vai) -- SB rattles
 * zhiishiigwan (vti) -- SB rattles ST
 * zhiishiigwanan (vta) -- SB rattles SB ==> this too is unglossed
 * to which the (vai) transformed into a noun means "rattlesnake" and (vti) transformed into a noun is a "rattle". The etmology for the root would be /šîhšîy/(gourd/rattle) + /kwä/ (transitivity suffix).  The word for "gourd" as "zhiishii(-g)" don't exist in Ojibwe (it is instead "okanakosimaan"), but in Arapaho the word for "gourd" is "siisiiy (pl. siisiyono')" and the verb word for "rattle" is "siisiiy", and etmologically have the same origins as the Proto-Algonquian version of the Ojibwe etmological root /šîhšîy/.  I'm at a dead-end on how to have an appropriate template, other than there needs to be nesting.  Also, to eliminate the long string of prefixes and suffixes, there ought to be a separate page of nothing but generic conjugation.  This is the approach Dr. Rand Valentine is doing on his Ojibwe page; Maliseet-Passamaquoddy electronic dictionary project is doing this as well.  For the Ojibwe, I would still have the Fiero spelling be the main spelling, but I would also provide two additional spellings: Syllabics of the pre-glyph glide, Eastern mixed-finals, independent sigma-form liquids (both pointed and unpointed), and Great Lakes Aboriginal Syllabics (both full and slacked).  Though the Eastern a-finals are the common form, I suggest the Eastern mixed-finals only because it shows the roots better.  I would need to consult with others with how to deal the scope.  For words common across all non-syncope Freelang Ojibwe dialects would not be a problem.  The problem would be how to present the dialect specific words and how to show syncope.  In the case of Odawa or Potawatomi (both well documented), if a word is given in its syncope form, one would need to provide all the different permutations of while if a syncope elimination rule tool/guide is given then they could go to any Ojibwe word and strip off necessary vowels, and if certain unacceptable combinations occur, those unacceptable combinations are further reduced.  However, this means it requires the user to know the full form.  I don't know who to ask for what to do for entering a syncope form and finding the full form.  For the Freelang, we had to look at the permutations of the word to restore the syncope-stripped vowel and syllables.  As for other Algonquian languages, they are either extinct or in critical danger of becoming extinct.  Scope would not be an issue to them.  Cree-group would be the only other Algonquian language where scope may become an issue.  So what templates are needed?
 * templates of grammatical tags common to Algonquian languages (this should be relatively easy)
 * template of wiktionary Algonquian word presentation (whatever that may be), incorporating templates for grammatical tags, images from the wikicommons, links to grammatical charts, and in the case of Odawa and Potawatomi, a syncope reduction guide and a auto-wild-card lookup tool to narrow possible choices (such that "Nishnaabemwin" would automatically treat the word as "ni-š?-na-pe-m?-hi-n?" and pull up all words fitting this pattern, i.e. not make the fortis/lenis distinction, not make the long/short distinction, all syllables must end in a vowel, all glides and aspirates dropped, no vowels may begin a syllable so "h" used as a sort-dummy. At least if one were to do a search with syllabics this would be exactly how one would proceed).
 * In the case of other Algonquian languages, they don't have fortis/lenis values, but they do have ,  and <'> that roughly maps to the Ojibwe fortis/lenis pattern, so the auto-wild-card lookup example above would also work for other Algonquian languages as well.  Thinking about this, this auto wild-card lookup example would also work for romanized Japanese.  Does this sort-of answer your questions and what the possible scope may be for Algonquian languages? CJLippert 05:43, 27 January 2006 (UTC)


 * :-) Many thanks. The fact that so much of it is beyond me is not your fault.  I appreciate your grasp of Algonquian linguistics, and I would not want to inadvertently impose Indo-European language prejudices on this group of languages.  This does put you in a bit of a leadership position on this topic.  With a little luck you will find other knowledgeable people to get involved.  I very much believe in having first nations languages as a part of Wiktionary; they collectively represent a value-added that could go a long way toward making Wiktionary a unique site.  My apologies if my help is limited to mere encouragement, with the occasional suggestion about co-ordination.  Nevertheless I cannot overemphasize how valuable I consider your efforts. Eclecticology 17:56, 27 January 2006 (UTC)


 * The Wikipedia pages on the Ojibwe language, Ojibwe writing systems, and Ojibwe grammar offer some good, though by no means complete, overviews of some of the grammatical structures common in Algonquian languages, as does the page on polysynthesis. gives examples and descriptions of most of the important grammatical categories and structures of Algonquian languages, focusing on Ojibwe and Cree. These might help give you a better idea of the grammars of Algonquian languages, if you'd like.


 * Unfortunately, a lot of this more technical stuff goes over my head. I don't really understand how these templates would work, but I'll do what I can to help out, CJLippert. You're certainly right that the more IE-centric layout of Wiktionary doesn't fit well with many languages, the Algonquian languages among them. --Whimemsz 16:05, 28 January 2006 (UTC)


 * Thanks for joining in. I'm certainly not familiar with what arguments there may be to adopting the Fiero system as a standard, but being the one to introduce it here as the standard can be a big advantage.  Now that you have a rough idea of where you are going, it might be a good idea to fully develop a page with this system.  When that is done I may be able to better comment on how it fits in from a big picture perspective. Eclecticology 10:29, 29 January 2006 (UTC)


 * Well, I would say to use Syllabics and Fiero Spelling, much like using Hiragana and Hepburn Rōmaji in Japanese. CJLippert 22:14, 29 January 2006 (UTC)

Word-sorting within a page
What about this case, with "duck" and its variants?
 * zhiishiib {na} duck
 * shib {root}{na} duck, duck-like water-fowl indicator

I still would like to give the highest common form listed, then increase by suffix, giving intransitives priority over transitives and inanimates over animates and have that preference loop upon each iteration of the derivatives. What would make the most sense in sorting these... by strict alphabetical order within each of the two groups? Would grammatical suffixes have priority over words joined after the word? Would suffixes have priority over prefixes or the other way around?

I have pulled out all the "duck" entries out of the Freelang, eliminated the cross-lookups, and the yet-to-be-in-Freelang entries that the project team is still working on to put the entry in a Fiero spelling.

zhiishiib {na}	duck *Zhiishiib {na}	Duck Clan zhiishiibakik {na}	kettle (Manitoba Saulteaux) *zhiishiibakikoons {na}	tea pot, tea kettle zhiishiib-anwi {ni}	duckshot *zhiishiib-anwii-baashkizigan {ni}	shotgun *zhiishiib-anwiins {ni}	duckshot zhiishiibenh {na}	duck (Odawa) *zhiishiibe {na} (Northern Potawatomi)	duck zhiishiibens {na}	duck: little ~ zhiishiibiinh {na}	duckling (Odawa) zhiishiibiins {na}	duckling zhiishiibikojigan {na}	duck decoy zhiishiibishimon {ni} (Ontario Saulteaux)	duck dance Zhiishiibi-ziibi	Duck River (MB) *Zhiishiibi-ziibiing	Duck Bay (MB) Makade-zhiishiib {na}	Black Duck Clan makade-zhiishiib {na} (Wisconsin)	black duck aajigadeshib {na}	waterhen amikoshib {na}	crested merganser (lit: beaver-duck) amikoshib {na}	wood-duck andawishibe (vai)	go duck hunting bikwaakoshib {na}	kind of wild duck (lit: ball-duck) gaagaagishib {na}	cormorant (lit: raven-duck) gaagaagishib {na}	raven-duck gaagaagishibi-neyaashi      Indian Point [of Mille Lacs Lake](MN) Gaagaagishibi-neyaashi      Ravens Point [of Lake Winnibigoshish](MN) gaagaagishibi-neyaashiing   Black Duck Point [of Leech Lake](MN) Gaagaagishibi-ziibiins	Raven Creek (MN) gaagaagishibens {na}	little cormorant gashkideshib {na}	black duck (Oji-Cree) ginishtinookweshib {na}	kind of wild duck (lit: Cree-woman Duck) ginoogweyaweshib {na}	long-necked duck ininishib {na}	black duck (Mississauga) ininishib {na}	mallard duck ininishibens {na}	little mallard duck makadeshib {na}	black-duck mishishib {na}	eider duck naabeshib {na}	drake nandawishibe (vai)	hunt ducks *nendawishibaaning	Laura Lake (MN) (lit: at the [place for] duck hunting) noojishibe (vai)	hunting duck wakeyawishib {na}	butterball (Clangula albeola) zagataaganishib {na}	teal [bird]

Some words in the list have already been grouped such that the derived words are listed below their originator. However, answers to the above stated questions will determin how each of these will be ordered. As a note, all dictionaries out there use spelling by their appropriate alphabetical order, from the start of the word. The exception is Cuoq where under "CICIP" he lists all CICIP and -CIP words, and within this section he then lists the "duck" class of words alphabetically. Suggestion appreciated. CJLippert 23:42, 2 February 2006 (UTC)


 * I hope that you realize that I am trying to respond from a position of total ignorance. What you have above looks more like a page out of a thesaurus than a dictionary.  Many of these words should probably be on pages of their own where a fuller explanation can be given about how the word is fromed, or why a suffix rather than a prefix is used in certain circumstances.  For lists of words, as you might find in a "Derved terms" section, alphabetical order is the usual standard, but an other order that makes semantic sense would not be wrong.  Eclecticology 07:27, 3 February 2006 (UTC)


 * Semantically speaking, Algonquian words are constructed a lot like the way a hanzi/kanji/hanja/hanta characters are constructed. So, I know the traditional radical sorting, but how would Wiktionary sort them?  Would one want to have an index of nothing but these "Ojibwe Radicals" and then derive each combination much like one would for the hanzi characters?  However, unlike a hanzi where it appears identical or nearly identical in other languages, "Ojibwe Radical" would work only for the Ojibwe-group of languages, which means a similar page would need to be set up for the Sac-Fox-Kickapoo group, Cree group, etc., etc.
 * The other choice may be to give the main word entry and its definitions, then just simply list all the derived words and their appropriate roots without defining any of them, and instead have links to those derived words and the etmology of how the roots were combined to form the particular words.
 * So, if we go with defining just the main entry and only listing the derived words, we could end up with something like this:

==Algonquian word/root {Syllabic, then Roman}== ===Etymology=== {listing the pieces in order, beginning to end, with the heirarchy of full words, root words, and grammatical transforms (prefixes, reduplication, initial vowel changes; suffixes for adverbs, verbs,  nouns} ===Definition=== {tags being in this order: vii, vai, vti, vta, ni, na, av} {viip, vaip, vai+o, vti2, vti3, nid, na-pt, etc. variations would all be listed as one of the 7 identified tags, but with the table link taking the user to the appropriate table for viip, vaip, na-pt, etc.} {tags will be linked to define what they are in the  appropriate "grammar" pages} {after each tag would be a link to the appropriate conjugation or declension table, thus eliminating the need for listing out the possible variations due to grammatical uses} ===Variant=== {listed alphabetically all significant reductions  (such as "chi" of "gichi" in the case of Ojibwe) and a variational word with the same meaning and  same etmology but with a different realization of  the construct} ===Syncope=== ===Transformations=== {consonant stems - in the particular Algonquian languages' consonant order} {vowel stems - in the particular Algonquian languages' vowel order} ===Synonyms=== ===Other Roman orthographies=== ===References===
 * 1) [ [etym 1] ]
 * 2) [ [etym 2] ]
 * 3) [ [etym ...] ]
 * 1) ({tag})([ [conj/decl table] ]) definition 1
 * 2) derived words:
 * 3) * [ [derived word 1.1.1] ]
 * 4) ({tag})([ [conj/decl table] ]) definition 2
 * 5) derived words:
 * 6) * [ [derived word 1.2.1] ]
 * 1) * [ [derived word 1.2.1] ]
 * 1) [ [var 1] ]
 * 2) [ [var 2] ]
 * 3) [ [var ...] ]
 * 1) syncope 1
 * 1) [ [transform 1] ]
 * 1) [ [transform 2] ]
 * 1) [ [syn 1] ]
 * 2) [ [syn 2] ]
 * 3) [ [syn ...] ]
 * 1) orthog 1
 * 2) orthog 2
 * 3) orthog ...
 * 1) ref 1
 * 2) ref 2
 * 3) ref ...
 * So with structure, only "zhiishiib" would be on the page; all words containing the "zhiishiib" would be mearly listed on the page with appropriate links in alphabetical order, "shib" would be under the "Transforms" with a link to its own page. Any Proto-Algonquian etmology references would be found only with those pages that are roots. CJLippert 17:35, 3 February 2006 (UTC)


 * At this point I would suggest that you go ahead and write up zhiishiib as an alpha-test of these principles. If there are any problems they will more easily make themselves apparent there.  Theoretical structures outlined as above don't give bugs a chance to appear so they can be fixed. Eclecticology 21:01, 3 February 2006 (UTC)

alpha-test pages
I've put together a series of alpha test pages. They are zhiishii, zhiishiigwe, zhiishiib and shib. At home, my browser is IE, and it is not showing the I-finals properly; however, I can check how it looks on NS at work. Generally, NS seems to handle IPA and other characters better than IE. If this all seems to work, then I think I will extract words from the Freelang database, put it in MS Excel spreadsheet, separate the entries from two columns to multiple columns for each entry, then ask for a bot to assist me in uploading about 20,000 words. Once uploaded, I will ask friends to add in paper-copy dictionary Volume-Page-Column citations. (See example in zhiishiigwe under "Cuoq" and "Baraga".) Similarly, I would need to find a good place to stick in the reference to "Nichols & Nyholm" in the same format. Since N&N already uses the Fiero spelling, it wouldn't fit under "Other Roman orthographies," so where would be a good place to put the Page-Column citation? CJLippert 06:53, 5 February 2006 (UTC)


 * I have added hidden comments to zhiishiib about formats. In addition, most of the related words that you want linked should be wikified.  For now this should at least apply to the romanized words.  Probably also the syllabics, but not everybody's browsers will handle these.  I don't think that there's a big rush for figuiring out how to handle syllabics.  I'll try to get at your question about references tomorrow. Eclecticology 11:09, 5 February 2006 (UTC)
 * Those comments do make sense. Thanks.  For zhiishii and shib, the Category:oj:word-roots might make the most sense. For all other zhiishii-derived words Category:oj:verbs, with zhiishiigwe also having additional category of Category:oj:snakes or Category:oj:reptiles, while the zhiishiigwan would have additional categories of Category:oj:ceremonial_objects and Category:oj:musical_instruments.  I have noticed that for zhiishii, the template for Proto-Algonquian needs to be changed because it drops in zhiishii into Category:Cree language, and it shouldn't.  I will change the entry levels later.  Thanks for your input thus far.  CJLippert 14:40, 5 February 2006 (UTC)
 * I have sliced and diced the zhiishiib page. Also, replies to your some of your questions have been posed as hidden comments. CJLippert 20:38, 5 February 2006 (UTC)

I have created templates for your reference section. Thus putting ((RZ:Ojibwe)) will add all five items from your list in the references section. If instead you put you will have an editable list when you save. From there it should be an easy matter to remove any that might not be relevant to the particular entry. Specific references within the article could probably be simplified to something like Nichols, p.87 with the understanding that the reference is to the Nichols book in the References section.

One more little point about categories "Category:oj:Reptiles" should have a capital "R". This is structured with sorting in mind since the software will sort upper and lower case letters separately. Eclecticology 21:12, 5 February 2006 (UTC)


 * Oh, on the page numbers, I have already got requests from several linguists who have asked specifically for the "volume, page, column, entry" order. I guess this is like Nelson's Japanese-English Character Dictionary use of "Monoharashi" reference numbers.  Also, the big proponent of this "volume, page, column, entry" order are those folks already familiar with such referencing with works citing Baraga. CJLippert 21:24, 5 February 2006 (UTC)
 * How would I have Wiki automatically take diglyphs into consideration such that "a" and "aa" would be sorted separately and "z" and "zh would be sorted separately? CJLippert 21:19, 5 February 2006 (UTC)

Bot-loading to populate Wiktionary
So, I have posed a question on the Wiktionary Main Page Talk on how to batch load to populate Wiktionary. Suggestion was made for the use of the bot pagefromfile.py. How should the source file be arranged to allow for this bot to extract the information? How will we know if the Bot understands the proper parts of the language and if it is the main entry or a derived term? CJLippert 17:34, 12 February 2006 (UTC)


 * When using the Python tools to upload, you need several iterations of test runs. Once you have successfully 'bot imported 10 entries, you'll need to request that the bot flag is set on the bot account (e.g. User:CJLippertBot or User:ObjiweBot) by having a vote in the Beer parlour.  Once the vote shows community approval (usually 1 week) the request is forwarded to meta, where a Steward will set the flag on the account.


 * The file format can be in essence whatever you want it to be for the upload. See the top of this file] for the programmer's comments.  In a text file, you'd have a start marker, (one line) the article title (in wiki-bold) then the contents, followed by an end marker.


 * Note some important things please :
 * Sysops are instructed to block bots that have not been pre-approved on WT:BP.
 * Once an entry is created, it is much harder to 'bot correct, so no mass uploading should be performed while the entry contents are still in flux.
 * You may need to quickly delete mistakes if the 'bot does something you don't want it to. Joining the conversation on irc://irc.freenode.net#wiktionary (e.g. using chatzilla) will give you quick access to several sysops and 'bot programmers.  It is a pretty friendly crowd; not as brusque as I sometimes am.  Some there know more about the robot architecture than I do.
 * Since much of this duplicates the www.freelang.net copyright content, it should be made clearer that you are the copyright holder of this information being uploaded and that you have also granted www.freelang.net use of your copyrighted material, to avoid errant copyright violation accusations.


 * For example:

Anishinaabemowin

=
Start=========

Etymology

 * 1) ᐊᓂᐦᔑᓈᐯᒧ (anishinaabemo)
 * 2) -win (nominaliser suffix)

Inanimate noun
Anishinaabemowin (ᐊᓂᐦᔑᓈᐯᒧᐎᓐ)


 * 1) Indian(2) language
 * 2) The Anishinaabe language, which includes:
 * 3) *Algonquin language
 * 4) *Odawa language
 * 5) *Ojibwe language
 * 6) *Potawatomi language
 * 7) *Saulteaux language

Synonyms

 * 1) ᐴᑌᐙᑕᒦᒧᐎᓐ (boodewaadamiimowin)
 * 2) *Potawatomi language
 * 3) ᓇᐦᑲᐌᒧᐎᓐ (nakawemowin)
 * 4) *Saulteaux language
 * 5) ᐅᐋᐙᒧᐎᓐ (odaawaamowin)
 * 6) *Odawa language
 * 7) *Ottawa language
 * 8) ᐅᑎᔥᒀᑲᒦᒧᐎᓐ (odishkwaagamiimowin)
 * 9) *Algonquin language
 * 10) ᐅᒋᐺᒧᐎᓐ (ojibwemowin)
 * 11) *Ojibwe language
 * 12) *Chippewa language

Transformations

 * unaffected: ᐊᓂᐦᔑᓈᐯᒧᐎᓐ(anishinaabemowin)
 * unaffected with syncope: ᐊᓂᐦᔑᓈᐯᒧᐎᓇ(nishnaabemwin)
 * initial change: ᐁᓂᐦᔑᓈᐯᒧᐎᓐ (enishinaabemowin)


 * reduplication: ᐊᐦᔭᔑᓈᐯᒧᐎᓐ (ayanishinaabemowin)


 * initial change reduplication: ᐁᐦᔭᓂᐦᔑᓈᐯᒧᐎᓐ (eyanishinaabemowin)

Other Syllabic forms

 * Fully pointed mixed-finals: ᐊᓂᐦᔑᓈᐯᒧᐎᓐ
 * Fully-pointed Eastern A-finals: ᐊᓂᐦᔑᓈᐯᒧᐎᓐ
 * Common Eastern A-finals: ᐊᓂᔑᓈᐯᒧᐎᓐ
 * Unpointed Eastern A-finals: ᐊᓂᔑᓇᐯᒧᐎᓐ

Other Roman orthographies

 * Algonquin: anicinabemowin
 * Cuoq: anicinapemowin
 * Manitoba Saulteaux: anihšināpemowin
 * Ontario Saulteaux: anihshinaapemowin
 * Baraga:

Dictionary information

 * Baraga:
 * Cuoq:
 * EOCOD: 309a-4.0, 501a-12.08
 * CDMO: 010b-15.0, 200a-03.1, 221b-01.01

=
End=========


 * Lastly, I'd like to recommend again investigating the IRC channel. As you go through the process, the helpful people there can be, well, an enormous help.  The collective language expertise of the participants is enormous.  Hashing out multilingual nuances is much easier for them, as many of the regulars have dealt with these issues in their own languages before.  --Connel MacKenzie T C 22:06, 26 February 2006 (UTC)


 * Hoi,
 * When you have the data in an excell file or a comma seperated file or .. I may be able to read it into the software that I have that can convert the data into a big file to upload it to a WIKI. The great thing is, when I have it in this format, I will be able to upload it to other Wiktionaries as well (and have it in a format that will help to import it into WiktionaryZ) :) GerardM 22:39, 26 February 2006 (UTC)


 * Um, Gerard, what fields should be delimited in that format? --Connel MacKenzie T C 22:51, 2 March 2006 (UTC)

anishinaabemowin

 * Note: This conversation was moved from User talk:Connel MacKenzie and User talk:CJLippert to Wiktionary talk:About Algonquian languages on February 26 2006 ~18:00 UTC.

Hello,

I think there needs to be some compliance with Wiktionary, for the terms you are entering. Our entry layout specifies the headers desired on the English Wiktionary. Diverging too far from the recommendation means that much fewer people will be able to understand what you've written.

The point here is to share information. I think the complexity you are adding to the format is not necessary, and detracts from your entries.

Non-ASCII headers are absolutely a very big mistake.

I've edited the entry anishinaabemowin once so far. Most important in that edit was the removal of the extended-unicode characters that were used in a heading. When I have time, I'd like to revisit it, to bring it in line with normal Wiktionary headings. If that means we have to add one or two to WT:ELE just for Ojibwe, so much the better.

--Connel MacKenzie T C 03:16, 24 February 2006 (UTC)


 * Please see zhiishii on the reasoning (2 different Syllabics spellings, both romanized identically). However, if Anishinaabemowin were to be in roman, so should all Japanese and Chinese and all other languages using non-ASCII glyphs.  Anishinaabemowin is not written in roman letters by majority of fluent speakers... located in Canada.  Forcing it to be exclusively roman would automatically insult all Anishinaabe communities there.  (This is not an issue with Anishinaabe communities in the United States since Anishinaabemowin in the US is normally romanized.) Other than Anishinaabemowin, Anishininimowin (Oji-Cree), Nehiyaawemowin (Cree), Blackfoot, Inuktitut, Athapaskan, Beaver, Slavey, Carrier and Cherokee would have similar issues.  CJLippert 04:08, 26 February 2006 (UTC)


 * You misunderstand greatly. This is the English Wiktionary, where people speaking English wish to learn about Ojibwe.  It is not where Ojibwe speakers come to learn English.  That would be on the Ojibwe Wiktionary (is there one yet?)


 * Heading levels in particular must not be in non-ASCII characters! There is no possible justification for not telling your readers what they are looking at...and the only assumption one can make about someone reading the English Wiktionary is that they understand English.


 * The current state of affairs with Japanese and Chinese characters is very slowly being addressed. At the time the English Wiktionary was initialized, User:NanshuBot went quite crazy and uploaded many thousands of terms in an unacceptable format - but even that tries to convey information about the words and characters in English.


 * --Connel MacKenzie T C 04:29, 26 February 2006 (UTC)


 * To clarify further, I do not have a problem with moving anishinaabemowin to Anishinaabemowin or to ᐊᓂᐦᔑᓈᐯᒧᐎᓐ. I am talking about the content of the entry itself; it should explain to an English reader what is going on.  If there are further technical limitations preventing it from moving to ᐊᓂᐦᔑᓈᐯᒧᐎᓐ that I don't know about, please let me know.  But that won't change what I am saying about what is inside the entry, no matter where it ends up.  --Connel MacKenzie T C 04:47, 26 February 2006 (UTC)


 * Excuse me, but I don't think you understand. Unlike English, Anishinaabemowin is experiencing "Language Rivitalization", for many means that they are learning Anishinaabemowin as a second language, but at the same time for those who speak Anishinaabemowin for a first language, the speak either English or French as a co-first language, and needing resouces to further develop their vocabulary.  Since English is the Lingua Franca in the US as well as in all other Anishinaabemowin-speaking communities in Canada (with the exception of the communities in Quebec), English is still the appropriate vehicle.  In addition, of all other Algonquian languages, Anishinaabemowin is the closest to Proto-Algonquian, which means many linguists are interested in Anishinaabemowin as well, and in academics, English is, again, the principle mode of communication.  Consequently, look-ups using romanized Anishinaabemowin is the most appropriate because there are almost as many variations on orthography as there are communities, but the Fiero romanization is considered the international mode of communication.  However, listing words only by romanized Anishinaabemowin would highly insult the Anishinaabe in Canada.  CJLippert 18:30, 26 February 2006 (UTC)


 * You are incorrect; I do understand very clearly. All the previous statements I made were made without any knowledge of the existence of this page Wiktionary talk:About Algonquian languages, however that does not change what I said in any way.  This is the English Wiktionary, not the Ojibwe Wiktionary.  That would be http://oj.wiktionary.org/.  See meta:Help:How to start a new Wikipedia for further information on that topic.


 * The example you gave above for a sample layout did not mangle any ===headings=== with non-ASCII characters. What you did to the English entry anishinaabemowin here on the English Wiktionary does not match the layout example you specified above!


 * Note again that I am not telling you where the entry belongs, nor under what titles. Since Wiktionary aims to have all words in all languages, it would make sense to me to include each spelling variation as a separate entry.


 * Please review this edit since you seem to have not seen or studied it yet. The edits there were made with no knowledge of your propsed template (a couple sections above this conversation,) but they match your proposed template much closer than the entry you had there.  What I did was to make the sections of entry anishinaabemowin perhaps comprehensible to someone speaking English.  That is what you had in your sample/proposed layout above; an explanation in English of the Ojibwe term.


 * Please also note that I am not making an idle complaint. Your use of the non-standard headings is causing secondary maintenance tools to function incorrectly.  If you can make these initial entries match the layout you described above, most of the problems these have caused for the maintenance tools become moot.


 * But more importantly, the content of entries on the English Wiktionary need to convey information in English, no matter what the language is that the entry is describing.


 * --Connel MacKenzie T C 20:09, 26 February 2006 (UTC)


 * Then, how would one separate two very different words that have identical roman spelling but different syllabic spelling? For the case where the a word can fuction as a verb or a noun or any other parts of speech but have identical spelling, both by syllabics and by Roman, this approach you are suggesting do make sense, but that is not the root of the problem.  Please do not make assuptions that I have not reviewed the changes, since I have, which is why I have pointed you to zhiishii to ask you how to address what you are asking.  Any suggestions on how to address this?  Now again, please take a look at zhiishii on the 2 different Syllabics spellings, both romanized identically.  Until this is resolved, I would not even consider the bot loading of nearly 20,000 words.  (Thank you the bot-loading info. :) ).  Also, resolving this now would allow for similar possible hiccoughs to be avoided for other languages using syllabics.  Don't yell at me, because this issue do need discussion... and I really do appreciate the discussion. CJLippert 03:43, 27 February 2006 (UTC)
 * p.s. Thank you for better clarifying the conversation movement info. CJLippert 03:47, 27 February 2006 (UTC)


 * Maybe the real question is should there be one page with the most inclusive spelling of syllabics and the Fiero romanization, then listing out all the spelling variations of both, or should there be one each for every single syllabics and one for Fiero, and then having links to each other to make things be a cohesive unit. Either way, there is no way to differentiate the Ojibwe-English presentation, because no matter what route, it will be Ojibwe speakers and Ojibwe leaners looking at definitions in English due to the prevalence of English use.  Whatever solution here would then help with the Ojibwe-French Wiktionary.  Also, the bigger question, the economic question, is if the Wiktionary project willing to tie up resources to otherwise redundant entries because there are about 200 Anishinaabemowin-speaking communities, where each community is a Nation unto itself.  If computing memory storage space is treated as a limited commodity, I highly doubt this approch would work.  At least with Cherokee, this is also a relatively easy problem to solve since they are cohesive (relatively few de-centralized communities) and all within the United States.  On the other hand, Inuktitut would greatly benefit from this dicussion because instead of the Canada-United States issue and ~200 communities issue Anishinaabemowin faces, Inuktitut is spoken on two different continents, two differing alphabets (Danish Roman (in Canada, United States and Greenland) and Cyrillic), and three major syllabics grouping.   Whatever solution made here, after further tweeking, would allow Inuktitut to participate in the Wiktionary.  On the other hand, if computing memory space resource availability is a non-issue, let's go ahead and have a page in roman and a page in syllabics (possibly only four major types with all the other variations filed under one of the four) with links connecting all five.  This seems to be the approach Japanese has for their two Kana pages, their "Romaji" page and their main gloss page.  However, again, as a reminder, Japanese is a cohesive community with a standard grammar, standard standard orthography, and with limited dialects while Anishinaabemowin can only be broadly categorized into 10 dialects, but there is no standard orthography, there is no standard grammar, and due to the sheer limited number of speakers of this language, there is a great flux so an entry that is acceptable today may not be acceptable within the next 3-5 years.  Since no one here have a definitive "magic crystal ball" as resouces that can be referenced, all we can do is to provide as much flexibility as possible to reduce the amount of wasted computing memory space. CJLippert 17:36, 27 February 2006 (UTC)


 * I think all concerns about disk space should be ignored. Let's simply aim for the "correct" form of listing.  When compared to Wikicommons (the image file repository) I think we can safely rule out disk space as a limitation in the next 10-20 years, for text-based entries.


 * Due to other recent activities, this page keeps falling too far down on my "watched pages" for me to track. If you feel the conversation is being neglected, please tack small reminders on personal talk pages, for participants like me.  --Connel MacKenzie T C 22:49, 2 March 2006 (UTC)


 * My day job has been taking a lot of my time lately. Hopefully, I will be able to get back to Wiktionary in the near future. CJLippert 04:04, 27 September 2006 (UTC)