On page 21 of your book on Dagaare, you state that there are nine oral vowels in Dagaare and as you note in your response to our paper, on page 23 of your book you state that ‘even though /a/ is realised phonetically as a [-ATR] vowel it occurs both with the set of [+ATR] vowels and the set of [-ATR] vowels.’ The paper of ours that you refer to reports on data from 5 speakers and appears to be inconsistent with this position. Taking measurements for tongue root values at three points throughout the low vowel, including points where we would not expect to see significant coarticulatory effects from the following vowel, we interpret our data as motivating a tenth phonological vowel. The formant differences observed are larger than what we would expect from coarticulation alone. We certainly know that ATR systems can vary as to whether they have nine or ten contrastive vowels. The data we report on in this short paper suggest that at least the variety of Dagaare that we examined is indeed a ten-vowel system phonologically. The question of whether Dagaare has nine or ten vowels is an interesting one and will be resolved through data collection and analysis. It is unclear why raising this question, and contributing data to it, would be interpreted as a dishonest framing of the issue.

Thanks for your response but you still do not address my concern. Your article is boldly titled as Dagaare /a/ not being neutral to ATR. Which earlier work said it is neutral? My works beginning with the 1997 CSLI book and all the way to current joint work with Arto Anttila don't say this. Is there another author who said this? If not why frame your "short paper" title this way?

Dagaare is a 9 vowel language, phonologically. If you want to contest this show it with all the evidence, both traditional like minimal pair analysis and "modern", like your instrumental analysis...and challenge those of us who have worked on the language all our lives. But don't misrepresent our work in order to embellish your analysis.

We are assuming definitions of neutrality that are standard in the literature; as stated by Harry van der Hulst in his Encyclopedia entry to vowel harmony, for example, (and with Tangale as an example), "The non-advanced low vowel /a/ in Tangale misses a harmonic counterpart. This is what we call a neutral vowel." (van der Hulst, Harry. 2016. Vowel harmony. Linguistics: Oxford Research Encyclopedias. DOI: 10.1093/acrefore/9780199384655.013.38)

Well, this is a 2016 work that is not even on Dagaare. I didn't state this in my 1997 book, The Structure of Dagaare. What i found out is that variants of the low vowel occur in both sets of ATR harmony processes. And that's what your instrumental analysis appears to confirm and I thank you for that evidence. But you didn't find something new and contrary to the 9 vowel system. If you want to claim there is a 10th vowel you have to establish minimal pairs, not just look at instrumental variations. All the vowels also have variants depending on the phonetic positions they are found in. Would you say they are phonemes because of that?

Phonetic variants alone wouldn't establish that they are phonemes, but there are certainly differences cross-linguistically between gradient phonetic coarticulation and discrete phonologically derived variants. Only if all phonology were structure-preserving would we expect vowels to be phonologically invariant if they were not contrastive. Anyway, stay tuned for more on Dagaare -- definitely couldn't fit everything into a 2-page report! Great that this is turning into a discussion. Hopefully next time it can be started without accusations of dishonesty!

Hi Adams, I am now reading the posts and I am happy about the high level discussion with my friend Doug.

You wrote „you find the contestation is dishonest“. My comment is more to favor the content as - we as senior brothers- should not discourage our younger scholars! see you all very soon.@

Douglas, actually, there was ingrained misrepresentation, and this could have been intended or unintended. If intended it is dishonest. That's for a fact and you could simply just say it was unintended and then we can move on. But if you try to defend this misrepresentation I have no qualms for seeing what the authors of the article did as dishonest in trying to appropriate some "finding" by claiming that earlier researchers said something contrary when that was not the case.

Very many scholars do this kind of thing these days and we must stay vigilant!

Firmin, I don't even know what is high level about this and I really wish it were or that it eventually turns out to be. It is important that the young people you talk know the issues. I have already explained my comment about the dishonesty involved...so far they are not ready to see what they claim I wrote as an unintended misrepresentation, which leaves my charge standing.

As for the content bit, there is nothing to write home.about this paper by way of them finding something substantially new. They have only just used instrumental analysis to confirm what we established scholars of Dagaare have always maintained. Dagaare is a 9-language, with PHONETIC variants of /a/ appearing in both ATR types of harmonies (advanced and unadvanced/retracted). Until they can establish two phonemic /a/s through minimal pairs and other means there is nothing to.write home about their short, cryptic paper.

When people start to work on a new language it is important not to begin with reckless, wanna-be earth-shaking statements of little substance.

I look forward to reading their next set of papers on Dagaare, and I hope they will exercise circumspection in what they write.