Recording animal/plant names that are easy to say but difficult to write !

[murphyslab]

@botswanabugs

Do iNaturalists need to become phoneticists and be able to write down all the various clicks they hear, properly and write them down as the IPA. I feel I’m too old to learn to do this. The variety of clicks in languages and tonal patterns involving click sounds, would I think be forever beyond my ability, to record accurate in writing.

You are being ridiculous. You are the one suggesting that iNaturalist needs to take on the role of linguist and record the sounds and names of species in “rarely used and perhaps dying-out languages”. That is the role of a linguist. And YOU are the one suggesting that it be taken on by iNaturalist.

My suggestion was not to learn the IPA, but rather find a dictionary & copy an IPA pronunciation entry which exists in many dictionaries. If the dictionary doesn’t have that, then just use the language’s own orthography to update the iNaturalist lexicon. The point of my suggestion was about making feasible an idea which is in and of itself a kind of overreach.

@botswanabugs

I would not expect ALL inat plant names to be recorded as sounds but only those which are tough with funny click sounds which are really tough for mere mortals to write and are unwritten by the speakers.

The thing is that one needs a system which can be applied to a database & a user interface. When a when those are created, one needs to think about a full set of uses and how it would apply to all entries, not just a handful of ones for your personal use case. By planning and thinking through, that idea can be extended to other languages and use-cases, not just a patchwork for what you personally find difficult. If it’s just something that you personally find difficult, then you should come up with something for yourself.

Perhaps you should ask yourself, “who do I really have in mind when I’m making this feature request?”

Is it just yourself in mind here?

@botswanabugs

I am an English boy who has been married into a family of traditional doctors, sangomas and herbalists for 36 years and live in Central Botswana.

Or is it the people for whom the language is their heritage?

@botswanabugs

Many languages here have few speakers and even fewer in the younger generation who are losing touch with the natural history knowledge of their elders. […] Could iNaturalist be adapted to be more inclusive of people who rely more on speech than the written word ?

Your recent statements seem to be at odds with the initial request.

Now let’s think about who this really would be meant to help. If a person is navigating the iNaturalist website, I think that we can have the expectation that they are able to read in whatever language they use. They probably wouldn’t require an audio version of a word.

Maybe it would be enough to just write out the names in whatever the local language is. Then if they need to know how to pronounce it, they can look it up in a dictionary.

If you look in a Merriam-Webster dictionary, for instance, you’ll find a pronunciation guide. How M-W writes out its pronunciation is slightly different than the IPA, but it is equivalent and can be converted to the IPA in many cases. However most English dictionaries don’t use the IPA because pronunciation in English is not fixed and varies regionally.

More geographically restricted languages tend to not have that problem. The orthography of several languages, particularly those which were written down by linguists, tends to correspond more closely to the pronunciation. In other words, they tend to have phonemic orthography. So a reader will instantly know how to pronounce the word. And if not, many of those languages’’ dictionaries will use the IPA for the pronunciation guide, if not for the entirety of how they spell words.

A ǃXóõ Dictionary, for instance, is written with a phonemic orthography, so it’s usually possible to convert between the written language and pronunciation directly.

@botswanabugs

The problem is, do the || X ? ! / \ properly record the click sounds in the names. The click symbols are far more vaied than can be recorded with a standard typewriter and do not appear in Wiki as diacritics.

Here’s what I do for symbols: I copy the ones I need into a text file from which I can readily copy and paste.

Again, taking ǃXóõ as an example, one could go into the handful of ǃXóõ entries in the Wiktionary and copy the necessary symbols which are standard to the language. And it isn’t that crazy to learn the sound of a dozen or two new characters, even if one middle aged or later.

So maybe just write down the names of those species from Botswana in the local language. If you want to help preserve the sounds, make recordings and upload those to the Internet Archive.

[botswanabugs]

thanks ! You are making me think !

[fffffffff]

It’s often hard to understand even English from just phonetic writing, that’s why e.g. Google translate has audio in it, it’s much easier, also there’re many languages that are dying, so overall having an audio version of names somewhere on Internet is not a bad idea, because very fast it can become not a heritage, but forgotten piece of history, but ofc it’s a too big of a task for iNat, so it might be something different, maybe a different platform that could be linked the way Wiki pages are linked.

[botswanabugs]

This article highlights some of the problems of using standardised IPA symbols for clicks whch vary from language to language. https://www.researchgate.net/publication/339909628_Cua_language_of_Botswana_Resilience_or_a_longer_road_to_language_loss

[murphyslab]

No, it does not appear to highlight such an issue. I would suggest quoting the relevant passage.

Alternatively, perhaps you have linked the wrong article?

I’ll add that my above comment also noted that one can and should use the orthography of the native speakers of the language to input the name in the relevant language field. The point of the IPA would be to provide a reference and a computer generatable version of the name for wherever audio may be required.

[botswanabugs]

Sorry I may have linked the wrong article

Writing Khoisan: harmonized orthographies
for development of under-researched and marginalized
languages: the case of Cua, Kua, and Tsua dialect
continuum of Botswana
Andy Chebanne

Page 5
In discussions with Christopher Collins, one of the dedicated
Khoisanists linguists, various Khoisan linguists often report that are sounds that are difficult to distinguish using the IPA and it is difficult to recognize and distinguish them using the IPA symbols. It should be noted that part of why the IPA hasn’t developed better ways of distinguishing sounds in these languages is that there are few researchers who have worked on more than one Khoisan language who have a particular interest in phonetics or phonology. Also, contrasts found in languages such as !Xo˜o˜ and ǂHoan are not yet understood across the Khoisanist community as a whole. The other issue in Khoisan phonology is that, for instance, the G|ui uvular ejective click, e.g. !q’ differs from the N|uu uvular ejective click !q’ in terms of timing. In G|ui, the uvular release is near the click release, where as in N|uu it is long enough after the click release to be audible as a distinct stop. That is, even though the IPA transcriptions of these sounds is identical, one sounds like a single click burst and the other sounds like a click followed by an ejective. The sounds do not sound
the same as the articulatory components are the same but the timing is different and so the sounds are not the same. G|ui, !Xõõ, and a few other languages with small numbers of speakers appear contrast velar and uvular places of articulation in clicks. Most other Khoisan languages do not appear to contrast these places of articulation in clicks (see Miller et al. 2007, 2009; Collins and Namaseb 2011).

---

What Im suggesting is that if IPA has shortcomings in representing click sounds in Cua, and native speakers do not yet have an orthography, then audio records do have value and could be recorded using iNat to ensure that the correct species for the name is identified. Later, after the species is properly identified then the information could be transferred to the Endangered Language Project. I do think iNat has a vital role in ensuring correct species identification for a hard-to-write language. The unqualified citizen scientist or local speaker (but not writer) in the field, can take a photo and record an audio of the name, and leave it to the qualifed biologists and linguists elsewhere to provide a correct Latin name and local name in the proper orthography.

[murphyslab]

Wouldn’t it be simpler to directly record local names for each of the species elsewhere? I mean, you can show photos of the species to a native speaker and simply record several in serial. There isn’t any case that you’ve presented in which recording native names to iNaturalist somehow is an efficient workflow.

[earthknight]

Wouldn’t the proper place for such a recording be on the species description page, rather than the observation itself?

That would require a Wikipedia edit I suppose, but that could be scraped for the name recording and automatically linked to the observation rather than recorded along with the observation.

[upupa-epops]

It could also theoretically be added to the list of links under “More Info” beside the Wiki page there, but I’m not sure how many anyone notices those links.

[tiwane]

I brought it up with other staff members and while we think this is a really cool idea, it’s beyond the scope of iNaturalist so I’m going to close the request.