[00:04:14] *** Quits: nascent (6d2b4f15@gateway/web/freenode/ip.109.43.79.21) (Quit: Page closed) [04:00:57] *** Joins: drbean (~drbean@113.196.171.151) [09:54:49] welcome to new week full of GF possibilities! [10:07:30] Can't wait! [10:13:14] *** Joins: Kaarel (6d2b4f15@gateway/web/freenode/ip.109.43.79.21) [10:22:26] I was thinkig about yesterday's conversation, where Aarne said that the only evidence we have from Egyptians etc. is that they wrote, no recordings [10:22:53] but this Ancient Greek writing system I'd definitely count as evidence of spoken language :P [10:24:07] And if you're trying to use GF as a measure of language complexity, I'd imagine some of it is dependent on the writing system [10:24:30] How much detail of the phonetic level it shows etc. [10:25:12] but can't a language have a complex phonetics which isn't reflected in the orthography? [10:25:17] Just got this feeling from both Irish and Ancient Greek, some rules are a pain to write for GF just because the writing system reflects that [10:25:33] Yeah, it can, but it wouldn't be so complex in GF [10:26:16] exactly.. so GF is a bad measure of complexity :P or rather only a measure of complexity of a writing system [10:27:36] jstar: well, english would be one such language, i suppose? [10:27:49] sure [10:29:49] I don't really have an intuition how much would it affect, sure there are things in the syntax that would be complex or not regardless of the writing system, but at least these liaisons and other sound changes really seem to be different on different GF grammars [10:30:28] If they're all just in one module, maybe that could be excluded and just look at the rest, if one wanted to do some analysis on complexity :P [10:30:40] Well evariste you've been working on that, what do you think? [10:31:12] (3 days idle, maybe he's not here very regularly) [10:32:42] In a way, the written and spoken "variants" differs in alot of way, I just think that GF aims to model principally the former one. [10:34:08] Yeah, but it is forced to take into account different levels of detail depending on the orthography of the language [10:34:14] So of course it is bad if you are interested only in spoken language but I think you can get interesting data on the written variant [10:37:46] inariksit: what do you mean diffent level of detail? [10:37:59] No but I meant that if we're interested in finding out how the GF grammars of two languages reflect the complexities of those languages, but one of the languages has very elaborate orthography and other doesn't show much phonetic detail, the first would need more gf code to take that into account [10:38:08] evariste: Like what Hans is showing now [10:38:27] That for instance the orthography of the word changes according to other words [10:40:36] And there are of course cases where some language is phonetically more complex; like French has "le/la Cxxxx" but "l'Vxxxxx" [10:40:46] And Spanish just has all the articles el or la [10:42:17] all my Gf problems seem insignificant in comparison to Hans' [10:42:21] :D [10:42:33] if you're having GF problems I feel bad for you son... [10:43:33] :D [10:48:54] evariste: For example this phenomenon http://en.wikipedia.org/wiki/Lenition#Grammatical_lenition , I'm pretty sure I could find similar things in English or Finnish or whatever, but apparently Scottish Gaelic reflects that in the orthography [10:49:38] And it's not even a question of agreement, like gender (they have that too), just that which sound does the preceding word end with [11:03:12] Oh he skipped the problem slide :---D [11:26:12] inariksit: what I'd say is that a more complex orthography gives you a more complex (written) language, regardless of whether this complexity reflects a more complex phonetic or not [11:26:29] (everything else being equal) [11:28:19] like in French, the prefix 'in' turns into 'im' in front of some letters (b,p,m) although there is no difference of pronunciation [12:16:48] Just realized this is a good point: in a case of digraphia one should probably specify which writting system is used when measuring complexity on the written language (and probably for other measures al well) [12:29:54] Oh he skipped the problem slide :---D [12:37:01] (Oops I apparently pressed up+enter) [12:37:20] evariste: Yeah, what I was wondering is that in the whole GF grammar, do you think it's a big factor [12:41:13] I don't have any data on that but I guess that it could make a significant difference in morphology (fon instance the size of the paradigms) but much less in syntax [12:41:29] Yeah [13:02:54] *** Quits: Kaarel (6d2b4f15@gateway/web/freenode/ip.109.43.79.21) (Quit: Page closed) [15:41:06] evariste: schweet [15:41:23] :) [15:41:59] *** Joins: chru (~chru@ip-109-43-79-21.web.vodafone.de) [15:52:03] *** Quits: chru (~chru@ip-109-43-79-21.web.vodafone.de) (Read error: Connection reset by peer) [15:52:46] *** Joins: chru (~chru@ip-109-43-79-21.web.vodafone.de) [15:58:10] *** Quits: chru (~chru@ip-109-43-79-21.web.vodafone.de) (Ping timeout: 268 seconds) [15:59:18] se tunne ku asiakkaan it-tuki kohtelee ku idioottia. [15:59:22] *herpderp* [16:08:13] no ei se mittää, kai se maanantai o niilläki. [16:08:14] :P [16:08:21] in the meanwhile, java updates. jee. [16:08:46] sorry, wrong channel guys, please ignore this. [16:12:02] regarding Hans' talk today: is he going to provide the slides at some point? i'd be very interested in reading them. [16:14:48] Well *everyone* should give me their slides [16:14:51] *** Joins: chru (~chru@ip-109-43-79-21.web.vodafone.de) [16:15:05] I hope the gentle reminder email will help [16:15:55] okay, great. :) [16:16:14] in the meanwhile, i will continue testing my practice grammar for Finnish in emacs. [16:19:53] *** Quits: chru (~chru@ip-109-43-79-21.web.vodafone.de) (Ping timeout: 264 seconds) [16:20:56] inariksit: did you try emacs on Windows? [16:21:12] well or maybe halla has something to say about it? [16:22:56] jstar: it works decently with Windows XP SP3 compatibility. with Windows 7 it kinda bugs, or at least it did on my comp setup. [16:23:16] could be that it's a local issue (i've had to heavily modify my windows 7 setup). [16:39:17] inariksit: did you get emacs to work? [16:40:39] Ah, I forgot [16:40:48] The download got interrupted and then I had other things [16:41:23] I'd prefer now do the verb paradigms for Estonian, let me know if you try it, I can add the suggestion to the guide [16:43:27] i assume emacs, not Estonian, in which case yes, I'll keep you updated. [16:49:57] :D [16:49:59] Yeah [16:56:10] *** Joins: chru (~chru@ip-109-43-79-21.web.vodafone.de) [17:01:28] *** Quits: chru (~chru@ip-109-43-79-21.web.vodafone.de) (Ping timeout: 245 seconds) [17:04:45] halla: and have you tried the shell in emacs with some exotic characters? [17:04:56] e.g. the Hindi resource grammar [17:06:04] jstar: not yet, but it does handle umlauts better. [17:06:14] well that's a start [17:06:16] :) [17:07:08] yes, so i am very hopeful about hindi and kannada, for one. [17:10:15] *** Joins: chru (~chru@ip-109-43-79-21.web.vodafone.de) [17:24:05] is there a good way to turn a Numeral into an NP (at the API level)? for clauses like "the answer is 42" or "the maximum number of students is 30"... [17:42:01] maybe you go from Numeral -> Det and then Det -> NP [17:42:23] mkNP (mkDet ()) [17:42:28] where is your numeral [17:42:48] (I used the RGL source browser to search by types: http://www.grammaticalframework.org/lib/doc/browse/) [17:46:23] that looks good. thanks! [18:54:44] *** Quits: chru (~chru@ip-109-43-79-21.web.vodafone.de) (Quit: WeeChat 0.3.7)