|WikiProject South America / Brazil / Colombia / Venezuela||(Rated Start-class, Mid-importance)|
|WikiProject Languages||(Rated Start-class)|
find in ethnologue 
Considering that not only is "Macu" pejorative, but that the two Macus are the same name (the difference in stress is an inconsistent affectation), should we perhaps move this to Nadahup languages? kwami (talk) 00:20, 12 August 2009 (UTC)
According to Patience Epps (2005) Nadahup only design Nadeb, Daw and Hupde languages, then, this name is not applicable to Nukak and Kakwa languages. At present time, Epps named Nadahup (Maku) family only Nadëb (Kuyawi), Dâw, and Hup-Yuhup languages. She asserts that Hup and Yuhup share over 90% cognate basic vocabulary; Dâw share approximately 75% cognate vocabulary with Hup and Yuhup (see also Martins and Martins 1999: 254); for the Nadëb language (also known as Guariba), the percent of its vocabulary that it shares with Hup, Yuhup, and Dâw has been estimated at roughly 50% (cf. Martins and Martins 1999: 254). Instead, Martins and Martins (1999) propose that Kakua-Nukak share only 35% of their vocabularies cognate with Hup-Yuhup, but they note that “the lexical data on Kakua-Nukak are scanty and these figures are provisional” (1999: 254).
Kakua-Nukak relationship with the other Nadahup languages is in question because “it was not possible to discover rules of regular correspondence” among the words, although they appear to “share a certain resemblance” (Martins 2005: 331-41). Epps 2005 has attempted to reevaluate the claim that Kakua-Nukak is related to the Nadahup family.
Mondragón 1999 had calculated 42% cognate basic vocabulary Nadëb - Hudë and only 24% Hupdë - Nukak and 20% Nadëb - Nukak but 26% Nukak - Hoti and 25 % Nukak - Puinave and he discovers four rules of regular correspondence Nukak - Puinave.
It isn't possible to build a linguistic group "Vaupes" with Hup-Yuhup and Nukak-Kakwa. This is a geographic group only because Nukak-Kakwa is linguistically longer than Nadëb from Hup-Yuhup.
Then, if you accept the Epps 2005 hypothesis, Nukak-Kakwa is out of Nadahup familiy but if you accept the Martins hypothesis, Nukak-Kakwa is not in the linguistic group "Vaupés" with Hupdë and Yuhup too. In this case, Nukak-Kakwa is separated but in the same family. Thanks.
- I've reverted you again until we work this out. One problem is that you only change part of the article, so that it contradicts itself; another is that you present Epps as the only possibility. We should of course cover Epps, but not exclude other research.
- I do not have access to Epps, so I cannot fully evaluate what you're saying. What I do see is "This work is a reference grammar of Hup, a Nadahup (Makú) language spoken in the multilingual Vaupés region of northwest Amazonia." That is, for Epps it would appear that Makú and Nadahup are synonymous, whereas in your edits you have one as a subset of the other.
- Martins suggests that the reason Nadeb appears so divergent is that it is not part of the Vaupés Sprachbund, that the other languages have converged under Tucano influence. But they also say that "There are enough differences between [Nadeb] and the others to establish them [Nukak, Daw, & Hup] as a distinct group within the family." If Epps disputes that, fine, but we shouldn't then present Epps as the only possibility; you're basically saying Martins is wrong because they contradict Epps, which is not encyclopedic. kwami (talk) 06:47, 16 August 2009 (UTC)
- No. The problem is that Vaupes group is not a linguistic group. You make a big mistake if you put Nadëb longer than Nukak from Hupdë-Yuhup.
- All the scholars in agreement, unanimously, that Nukak-Kakwa (only 24%-35% probables cognates) is longer than Nadëb (42%-50% secure cognates) from Hupdë-Yuhup, consequently your diagram are mistaken, you can not make a group Vaupes with Nuakak and sudgroup Dâw-Hupdë in front a separated Nadëb. Instead of that, you can build a group with Nadëb and subgroup Dâw-Hupdë(+Yuhup) and a separate branch to Nukak-Kakwa.
- The controversy Eppps(2005) - Martins is about Nukak appurtenance: Epps(2005) puts out Nukak-Kakwa and includes in Nadahup (Maku) Family only Dâw-Hupdë(+Yuhup) and Nadëb, instead of Martins includes Nukak-Kakwa in Makú or Nadahup Family and others prefer to call Nadahup stock like subfamily or group of the Makú Family or named this Puinave-Makú Family or others named Macro-Puinavean but all scholars accord in a topic: Nukak-Kakwa are longest than Nadëb with regard to Dâw and Hupdë-Yuhup. Nadahup could be only a group or subfamily part of Makú or Puinave-Makú family intro Macro-Tukano (Greenberg) or a family into Macro-Puinavean (several scholars) or Nadahup has not conexions with Nukak-Maku and Kakwa-Maku.
- Please, it's necessary to change the diagram and infobox so as to present real distances from Nuakak to the rest of family and it's better to modify the lit because identity Nadahup = Makú is only a hypothesis or rather a synonyms proposal. I had write these three changes in the last edition (only one before). Thanks,
- Hhmb (Hhmb (talk) 16:23, 16 August 2009 (UTC))
- Sorry, you are simply wrong re. the position of Nukak. There is no such unanimity.
- The diagram does need correction, but not in the way you say. There are two classifications proposed by Martins (1999): two branches, Nadeb and Vaupes, and three branches, Nadeb, Nukak, and Hup-Daw. In neither case do they suggest that Nukak is more divergent than Nadeb, despite the lower %age of cognates, flatly contradicting Epps. (Pure lexicostatistic %age of cognates is not a reliable way to classify languages, so this does not prove Epps is correct, as you seem to believe.) To accommodate Epps, we would need a third diagram. The only way to synthesize the two would be to assume that Martins got confused, and that wherever they wrote "Nd-Ku" (Nadeb-Kuyawi) they really meant "Kk-Nk" (Kakua-Nukak), and that wherever they wrote "Kk-Nk", they really meant "Nd-Ku". If that were the case, we should be able to find a mea culpa published somewhere. Pending that, we simply cannot say that scholars are "unanimous", and we should not replace Martins with Epps, especially without noting that this is the opinion of a single scholar and is contradicted by other scholars. kwami (talk) 23:53, 16 August 2009 (UTC)