Talk:Protea canaliculata

Authority

There's some confusion over the correct authority for the name Protea canaliculata. Right now, IPNI has two entries, neither with a date, but this one is the oldest. Taxonomic Literature, usually taken as definitive in these matters, gives the date of Plate 437 as December 1806, and the author of the text of Volume 7 in which it occurs as George Jackson (it's not clear whether the preceding "probably" applies only the first name in the list or to all three) – see here. I'll contact IPNI for clarification, but for now it seems best to leave the authority in the taxobox as "Andrews" as per IPNI. Peter coxhead (talk) 07:54, 23 July 2020 (UTC)

I had a reply from IPNI, referencing the ICNafp, Article 46. Art. 46.8 applies: "In determining the correct author citation, only internal evidence in the publication as a whole (as defined in Art. 37.5) where the name was validly published is to be accepted, including ascription of the name, statements in the introduction, title, or acknowledgements, and typographical or stylistic distinctions in the text." So as there's mention of the author of the text in The botanist's register, the name must simply be assigned to Andrews. IPNI will be updated with the date as per TL. Peter coxhead (talk) 14:44, 23 July 2020 (UTC)

You're watching like a hawk! But I have a good reference for the authorship attribution to Haworth, sorry, got distracted with another Protea before posting it -so much info, was thinking about how to summarise it nicely. You better read it yourself before making a decision here. To wit, it is certain Haworth published all the names in The botanist's register in that edition (actually, a series of individual articles later bundled) although authorship was rarely given in that magazine, Andrews quit as editor in 1805, a few years before. Salisbury also attributes Haworth for this name in a publication soon afterward, as did Haworth himself for other names in this edition (i.e. Agave americana, long story). The Flora Capensis (1915 abouts) attributed Haw., quoting my source. The source uses ephemera such as the text on the wrappings around the issues (still preserved in a few libraries!) and advertisements in other papers. I would like to take a close look at the other major African floras regarding that area, but I believe this is simply a case of bad database info inherited from the IK. Bear with me a bit while I gather the documents. Leo Breman (talk) 19:47, 23 July 2020 (UTC)

There are two main issues here.

1. Who is right about who wrote the text, TL (usually treated as the 'bible' for such taxonomic matters) or other sources? Either way, all must be given in the article to retain WP:NPOV. The relevant part of TL is old, so I can well believe that it's not correct, and we could say something like "however, newer sources ..."
2. What is the authority for the name? This is entirely governed by the ICNafp, which is absolutely clear that unless there is internal evidence to the contrary, the stated author of the work must be treated as the authority. So unless The botanist's register somewhere says otherwise, the name must be asigned to Andrews, because this is the name on the titleplate, regardless of who actually wrote it. There can be no dispute over this. It's irrelevant what Salisbury or Haworth said elsewhere, as per the quote I gave above from the ICNafp. This is not a case of bad database info inherited from the IK (although there is undoubtedly still a lot of that in IPNI).

Peter coxhead (talk) 08:15, 24 July 2020 (UTC)

Well, Peter, please read this 1916 article. Andrews is cited in earlier 19th century works such as the Genera plantarum or Prodomus, hence the writing of this article. It seems the local floras switched the attribution to Haworth after 1916 (looking for a copy of Flowering Plants of South Africa, 1925, for extra verification), but the Index Kewensis never bothered, so a local monographic flora is now in dispute with an old database...

Bible, shmible. People make mistakes, especially post-docs stuck on data-entry duties.

Regarding your second issue, should this rule be applied consistently, then a large percentage of the Protea names have the wrong authorship attribution, see anything with the attribution 'Salisb. ex. Knight'. Following the established practice of the IPNI, then 'Haw. ex. Andrews' would be more correct. In this case, the work was written and published by Haworth, which was made clear at the time in the accompanying documentation, but authorship is not stated within the article. The title plate is rather misleading, the book at the BHL is actually a special (reprint?) edition which bundled all the (bi?-)monthly issues of this magazine into a series of books, and recycled the title plate for each book in the series. I.e., it is not actually the title plate to this work! ICNafp doesn't actually apply in this case, as there is no internal evidence at all in the original publication regarding authorship or even date, so even attribution to Andrews is relying on external evidence, in this case, the later addition of a frontispice. And where ICNafp does apply regarding the uncredited work of Salisbury in the book by Knight, the IPNI is not following the rules!

Obviously, the authorship issues should be explained in the taxonomy section of this article. As for attribution in the taxobox, maybe a "see taxonomy"-type thing would be best.

Regards, Leo Breman (talk) 09:31, 24 July 2020 (UTC)

Interesting. However, it doesn't change the formal botanical authority, which must follow the ICNafp, and so far there's no evidence that Kanchi Gandhi at IPNI was wrong in rejecting my suggestion that it should be "X in H.C.Andrews" where X is either "Jacks." as per TL or "Haw." as per other sources. (It can't be "ex" because "ex" requires explicit reference by Andrews to another person's first use of the name, as per the ICNafp again.) If it's indeed that case that different authorship was made clear at the time in the accompanying documentation, then let's supply that documentation to IPNI and see what they say. (However, unless it's online, with Covid-19 restrictions they probably can't access the Kew copy of the Repository.) Until then, it's reasonable to use the authorship of the work as a whole, which clearly originated with Andrews, whoever took it over later.

And yes, there are thousands of errors in IPNI. At one time I kept a list of the errors I have reported and which have been corrected – almost every time I have seriously worked on a plant genus I have found multiple errors – easier now that BHL has so many old sources online. One problem (apologies if you know this) is that IK recorded mentions in the taxonomic literature, rather than just protologues. IPNI simply doesn't have the resources to do much pro-active checking, if any, although I find them very prompt to respond to my queries. I'm still caught out by the arcane rules of the ICNafp, however! Peter coxhead (talk) 10:07, 24 July 2020 (UTC)

Comments: Yes, I noticed the usage of the word ex was incorrect regarding Protea species, it's used to validate nomina nuda as far as I know, but that is how it is being used here regarding Salisbury ex. Knight. Per TL, the frontispiece was printed in 1797 as the second issue of the magazine, for use in binding the first volume. There were originally title pages, but these are no longer extant. The frontispiece was simply reused, so I can't see how ICNafp actually applies here: only the original publication is relevant. Regarding the accompanying documentation, the source I used says that the wrappings around the issues were preserved in the Banksian library, obviously we don't have online access to that. I noticed the IPNI updated their website with new dates per TL. Something bugging me now is the discrepancies between the 1916 source (Britten) with the TL, like the dates (TL claims most of the plates are dated, this is patently not true regarding the volume we are discussing). TL provides no evidence, but I will try to read their sources: the German one by Harms looks promising. Leo Breman (talk) 15:42, 24 July 2020 (UTC)

The table pasted into this copy is interesting. I would put the possible dates, 1806 or 1807, in the article. Peter coxhead (talk) 19:37, 25 July 2020 (UTC)

Christies note of a copy they sold [1] (for USD11,250) that some of the plates were "shaved". I wonder if dates were originally present or are present in some editions/copies? Not something we're likely to find out. Peter coxhead (talk) 19:56, 25 July 2020 (UTC)

I found this 1941 article by Nakai on dates. He says 1806 for this plate (437), oddly enough specifically without December, as in TL. I now see I misread Britten in dates, and if it is correct in that issues were published as a set of five species, December 1806 is indeed most logical. Harms concurs in date I gather from the link you've provided above. So this problem might be sorted. I'm writing up a draft with all the relevant opinions on authorship in chronological order, with all 12 taxonomic references we've gathered so far. Just names, dates, books. That should be the most neutral approach. All this nomenclature might be boring for the average reader, so I've already begun to expand the other aspects of this species. Leo Breman (talk) 22:08, 25 July 2020 (UTC)