Early views are temporary papers, so the Code does not accept them as published (cf. Arts. 9.9 and 21.8.3 in the Amendment to the Code, 2012). Consequently the year to be cited is only the one in which the definitive version is placed online.

So in fact, Early View equals presentations at conferences and unpublished manuscripts. Potentially dangerous, thus.

A reason behind _not_ referring to as-yet-unpublished taxa in other papers is that unforeseen events may occur that delay the publication of the descriptive work, in which case your paper (discussing but not describing the new name) could come out beforehand, resulting in a nomen nudum authored by _you_, not the descriptive author.

You have to distinguish "Early View" from "Accepted author version posted online". Please read the International Code of Nomenclature for

algae, fungi, and plants (Melbourne Code) (http://www.iapt-taxon.org/nomen/main.php?page=title), art. 30, Ex. 6:

Thank you all for your input! I was mainly worried about the nomen nudum trap (potentially by myself), and as I understand, an Early View considered final by the publisher (version of record) is considered valid according to the code, but the preliminary pagination has to be replaced with the final in due time (but the preliminary pagination page may be added between brackets, preceding the final pagination number.

As you see, I am talking about plants (you know there's a different Code for animals, etc.)... In the Code there are no enough examples about "Early View" publications, but I suppose that, in the Example 6 that I mentioned, before the publication of the printed version (on 27 February 2012), the correct name should be:

Partitatheca D. Edwards & al. in Bot. J. Linn. Soc. (online, DOI: 10.1111/j.1095-8339.2011.01207.x): [2 of 29]. 1 Feb 2012. Or something like this...

For ICZN, see http://iczn.org/content/electronic-publication-made-available-amendment-code

For pdf, see also http://www.mapress.com/zootaxa/list/2012/3450.html

ICZN (2012) Amendment of Articles 8, 9, 10, 21 and 78 of the International Code of Zoological Nomenclature to expand and refine methods of publication. Zootaxa 3450: 1–7.

This really makes mess of publication dates. Perhaps taxonomic articles should include a line below the abstract stating the date of publication of the "version of record" As it stands, I believe that the "early view" date disappears in journals like Cladistics and Systematics and Biodiversity when the printed version of the article appears. Even from the electronic downloadable versions of the articles.

A validade de um táxon é a data de sua publicação.

The validity of a taxon is the date of its publication.

The validity of a taxon is the date of its publication.

The ICN is actually confusing regarding this issue. Art. 41.5 states that combinations are only validly published (after 1953) if a full reference to the basionym is given, including a page or plate number. While this is not directly related to valid publication of a new taxon, it requires that the protologue contains a final pagination. However, with early view or online first, usually no final page numbers are given. This then poses a problem if someone later wants to recombine a taxon into a different genus, for example. Art. 30.3 suggests that if no final pagination is present in early view or online first versions, then the pagination can be cited indirectly as [2 of 29] in the example. The same article also states that once the pagination is known, the actual page can be cited [230], but with the date of the early view / online first post (where that pagination was not yet known). While this example makes early view / online first publications valid, it is a contradiction in itself, as it mixes information from two technically separate publications with different publication dates. If you have to cite the original publication date in a paper [1 Feb 2012 in the example], it makes no sense to at the same time cite the final pagination published in a later version [230] as part of the early view publication, as suggested by the ICN, because they are NOT both part of a single version of the document. Correct would be to cite both versions separately, so as to avoid confusion when someone checks either protologue and does not find the final pagination [230] in the early view version [1 Feb 2012].

In short, online first and, under certain circumstances, early view publications are valid, but the situation is rather messy. Especially with journals where there is a lengthy delay between online first and paginated versions (sometimes a year or more) or journals such as PLOSOne which have abandoned sequential pagination all together and officially cite only an e-number (while each article starts automatically with page 1, even in the same issue). So for example the typical citation for a name, saying "PLOSOne 7(12): 1 (2012)" (using the internal pagination) could refer to any paper published in issue 7(12) and hence, according to ICN Art. 41.5, cannot be considered a direct reference anymore. On the other hand, the citation "PlosOne 7(12): e51392 (2012)", which is the official citation, would be invalid for a combination as "e51392" is not a page number but refers to an entire article.

Any taxonomist should strictly follow the clear ICZN rules for electronic/digital publication. I observe many cases where there is no reference (number) in the pdf versions to the ZooBank, where taxa have to be registered. Without such registration, the taxon does not become valid before the print version is available. So - perhaps never.

Update to my previous post: at least some journals are considering the "early view" publication date as the publication date of reference for nomenclatural acts. This date should appear in the printed version of the article, although sometimes it does not. Sadly, this choice seems that this suggests the decreasing relevance of paper publications.

some journals use the early view of their articles to attract the attention of specialists and consider any proper objections to avoid further complications. I think as long as any digitally released article has no valid publication date approved by the journal, there is no need to include the newly described taxa as a valid one.  

As I have noted earlier, digital/online taxa are only valid if they are formally registered with ZooBank. The pdf has to show in a prominent place the ZooBank registration number - which is not too difficult to obtain. If the number is there, then the taxon is valid as soon as the journal is formally published online. An even earlier private distribution of a pdf, before the journal/publisher homepage shows the manuscript, is no publication at all. If there is no ZooBank registration, the taxon is invalid until the paper version is printed. If there is no paper version of a journal at all, the taxon remains invalid. But a taxon may become valid when the ZooBank registration is eventually done later - but this requires the publication of the registration number in some way. In this case I am not too sure and the ICZN and ZooBank should be asked.

In any case: ALL GOOD JOURNALS DO THE ZOOBANK REGISTRATION - and should advice their authors to do it. 

I completely agree to Ralph Becker. And a name is indisponible until the ZooBank number have been published.