Wikipedia talk:Featured article candidates/Metalloid/archive1

Comments by Squeamish Ossifrage

Entire section moved from the review page due to length, number of fundamental concerns, and the article's primary editor's responsiveness, or lack thereof. Squeamish Ossifrage (talk) 15:15, 27 March 2013 (UTC)

• Oppose on multiple grounds. Squeamish Ossifrage (talk) 17:13, 12 March 2013 (UTC)

1a: One significant concern regarding the well-written criterion is that much of the article consists of bulletted lists rather than prose, even when the information being conveyed is not inherently suited to the list format, to the extent that I do not believe this meets the Good Article criterion 1b regarding the use of embedded lists, much less the FA prose standard. The "Definitions" section has this problem, as does the "Biological interactions" section (albeit in a different sort of format), and arguably the "Derivation and other names" section. The article's treatment of astatine consists of a numbered list formatted into the appearance of prose. I didn't spend much time examining the prose quality of specific passages because I think the broader problems need resolved first, but I suspect there's considerable room for improvement there as well.

1c: Ideally, we would cite comments about the frequency with which elements are referred to as metalloids to reliable sources that discuss that frequency by reviewing the extant literature themselves. I'm not sure if that is available in an easy package. With careful citation and wording, some claims about frequency-of-use are probably possible even without this sort of metacitation while still avoiding original research. However, this article relies heavily on List of metalloid lists, another Wikipedia article, for its determinations of citation frequency. I understand the motivation here, but Wikipedia articles cannot use other Wikipedia articles as sources, especially for such fundamental underpinnings of much of the article structure, and especially at the FA level.

2b: I'm pretty dubious about some of this structure as well. The article seems to spend much less time discussing the group as a group, versus examination of the individual members. It's possible that's mandatory given the poor specificity of the group itself, but the order of content also feels disjointed. We begin with a broad examination of the group as a whole, then spend the bulk of the article on the individual elements, then return to category-level material. Sections like "Quantitative description" and "Nomenclature and history" feel like they should be more closely associated with the sections on "Definitions" and "Properties of metalloids", rather than at opposite ends of the article. Likewise, I suspect that "Typical applications" / "Biological interactions" / "Abundance" should be construed so that they discuss the whole of the category, rather than existing as subcategories of the "Elements commonly recognised as metalloids" section. The discussion of allotropes takes place in a subsection of "Elements less commonly recognised as metalloids" but mentions arsenic and antimony, which are discussed elsewhere. As a separate concern, there's a huge amount of information packed into the notes, some of which almost certainly shouldn't be relegated to the basement. What is currently note 38, in particular, strikes me as not being the sort of thing notes are intended for.

2c: I'm not going to quibble with the nonstandard citation format on its own merits, although I suspect others might (in particular, the lack of a concluding period seems odd to my eye), but there's a lot of formatting inconsistencies. Many sources have publication locations, but some do not (e.g. Aldridge 1998). There is sometimes, but not always, a comma before reference numbers (e.g Barfuß et al. 1981), and some reference numbers are missing (e.g Allen & Ordway 1968) or not correctly linked (e.g Askeland, Fulay & Wright 2011). The sort of chained-reference format used for Eberle 1995 and Keller 1985 probably isn't acceptable at the FA level. What makes Allenden 2012 a reliable source, and what is its year of publication? What makes Brown 2007 a reliable source? Gray 2010? Hoppe 2011? Tuthill 2011? What is Haller 2006? It's hosted by the Office of Scientific and Technical Information but the reference doesn't cite any actual publication information, and I don't see any. Why are you using Hawkes 1999 at all, especially over Hawkes 2010, which you also cite (including citing both for the same claim at one point). In fact, several times, there are citations to multiple editions of the same work (e.g Metcalfe, Williams & Castka 1966 / 1974 / 1982); if this isn't necessary, it feels like WP:BOMBARDMENT. There may be more citation and referencing issues as well. I started to skim about halfway down the second column.

Valuable feedback, thank you. Briefly, as I understand it, you have concerns about: 1. Lists; 2: Frequency citations (relying on list of metalloid lists); 3: Structure; 4: Notes; and 5: Formatting inconsistencies. Upon first consideration I don't see anything insurmountable about addressing any of these issues. However at least one (#2) may result in some interesting discussion (i.e. I'm not sure I see an issue there) but let's see how things pan out. I'll respond further, shortly(ish). Thanks again. Sandbh (talk) 12:25, 13 March 2013 (UTC)

The problem with list of metalloid lists is maybe the most serious flaw the article has. That list has assembled a huge volume of references, and then counts through them (sometimes awarding half-points) to determine the degree to which there is consensus for including one element or another in the metalloid category. That process is then the basis for much of the fundamental structure of this article. But our policy on academic consensus says that this sort of effort is novel synthesis. And that sets aside the separate issue that one Wikipedia page cannot rely on another as a source. All these claims that one element or another is commonly considered a metalloid, or uncommonly, or rarely ... all of them need a reliable third-party source to make those claims for us. Otherwise, a fundamentally different approach to organizing and conveying this information and the disputed nature of the category is required. Squeamish Ossifrage (talk) 16:33, 13 March 2013 (UTC)

1. Bulleted lists.

• Could I check your statement that 'much of the article consists of bulletted lists.' The main body runs to about 8,300 words. Around 858 of those words are in bulleted lists. So, around ten per cent of the article consists of bulleted lists. That does not seem like much.
• Could I also request clarification of your concern that the article conveys information that is 'not inherently suited to the list format'. As per your comments, and as an example, the defintions section includes four definitions/extracts (quoted verbatim from the literature), and presented as a bulleted list. I'm perplexed as to which aspects of this information aren't inherently suited to list format.

Please look at Wikipedia:Embedded lists. Much of the material presented as lists here does not need to be presented as lists, as prose is generally preferred where both are reasonable. That's especially true at FAC, where the prose is expected to be "engaging, even brilliant, and of a professional standard". I haven't really looked at the other prose, because if 10% of the article does not meet that standard, then the article does not meet that standard. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

I think what you are saying is that the lists in the article are less reasonable than what their prose equivalents would look like. Therefore, since prose is preferred, they should be rendered in prose. I agree with this position, in principle. My objection is that, for example, the first list—which includes four metalloid definitions/extracts (quoted verbatim from the literature)—is a better way of presenting this information than prose. If you think such a list would look better as prose I'd be grateful for any suggestions as to how this might be accomplished. (When I was working on the pre-FAC feedback for this article, this list was one of the few I thought would be more appropriately retained in list form.) Sandbh (talk) 07:52, 19 March 2013 (UTC)

2. Frequency citations (relying on list of metalloid lists).

• I agree with the Wikipedia guideline that a Wikipedia article cannot use another as a source. In this case, the reference to list of metalloids lists avoids the need to add the list of metalloids lists and its 194 references to the metalloid article. I submit this is a better outcome for Wikipedia and, if needs be, can be regarded as a reasonable exception to the guideline in question.
• 'List of metalloid lists' is not OR or novel synthesis, as I understand it. It does not advance a position. It is consistent with the seven references given at citation 49 and footnote 8 that 'Boron, silicon, germanium, arsenic, antimony and tellurium are commonly described as metalloids.' It is consistent with citation 50, which notes that, 'One or more from among selenium, polonium or astatine are sometimes added to the list.' It is consistent with Rochow's observation (1966) that, 'For one reason or another, it is sometimes desirable to think of a few other neighboring elements as metalloids'. He goes on to list carbon, phosphorus, and bismuth as examples. It is consistent with Hawkes (2001) who wrote that, 'Lists of “semimetallic” elements differ. Most include all the elements (other than aluminum and beryllium) that are next to the zigzag line separating metals from nonmetals on the periodic table. Most exclude bismuth and selenium, and most exclude either polonium or astatine or neither but not both.'

Adding 194 references to this article and counting them here instead of in list of metalloids lists would admittedly solve the problem that you are relying on another Wikipedia article as a de facto source, but it wouldn't fix the fundamental problem that this isn't the way you should be handling the problem to begin with. The article needs to cite sources that themselves talk about how often given elements are likely to be considered metalloids, rather than having you make that determination. You may not be advancing a position with this analysis, but you do "imply a conclusion not explicitly stated by any of the sources". Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

The situation is complicated by the absence of agreement in the literature as to which elements are appropriately classified as metalloids.

As noted, the article cites seven references to boron, silicon, germanium, arsenic, antimony and tellurium being commonly recognised as metalloids. Hawkes (2001) further notes that 'lists of "semimetallic" elements differ' and that 'Most exclude bismuth and selenium, and most exclude either polonium or astatine or neither but not both.' Rochow (1966) had earlier said that for one reason or another, it was sometimes desirable to think of neighbouring elements such as carbon, phosphorus and bismuth as metalloids (he later mentions P, Bi and Se). Then there are the outriders: elements not specifically mentioned by either Rochow or Hawkes, but for which isolated references to their categorisation as metalloids can nevertheless be found in the literature (as given in the list of metalloid lists article).

On the basis of the above I think the literature is reasonably clear that there are: six elements commonly recognised as metalloids; two elements which don't appear quite as often (polonium and astatine); and several elements either sometimes classified as metalloids, or which are self-evidently categorised as metalloids only in isolated references.

Since there is no agreement as to which elements are metalloids, a point I make repeatedly in the article, and since this was the cause of the slow edit war the article used to be subject to, I compiled the list of metalloid lists (incomplete as it is) in order to see what could be discerned from the literature by simple artihmetic. The results were consistent with: six elements commonly recognised as metalloids: as well as with Hawkes (2001); and Rochow (1966), and the occurrence of outrider or isolated metalloids. There is no original research involved in counting the number of times a particular element appears as a metalloid in an incomplete lists of 194 lists of metalloids.

I then expressed the common threads of what can be found in the literature, in my own words, again noting there is no agreement as to which elements are appropriately classified as metalloids.

If there is original research here then maybe it is a technical breach of WP:OR but it is certainly not a breach of the spirit of WP:OR. The article neither advances a position nor contains content inconsistent with mainstream thinking. Sandbh (talk) 12:58, 22 March 2013 (UTC)

3. Structure of article.

• The overall structure is an outcome of previous reviews. The sketches of the elements commonly recognised as metalliods were added in response to feedback from a reviewer. This feedback was along lines that the ordinary reader would expect to see content about the individual metalloids, fairly early in the article.
• "Typical applications" / "Biological interactions" / "Abundance" are subcategories of the "Elements commonly recognised as metalloids" section. This is because the primary focus of the article is on the elements commonly recognised as metalloids.

The article's primary focus should be "metalloids". Technological applications, biological interactions, and terrestrial abundance are all examples of the types of information that can and should apply to the category as a whole, and this article should discuss those subtopics in that context. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

I hesitate to discuss the technological applications, biological interactions, and terrestrial abundances of the 22 other elements less commonly to rarely classified as metalloids since the body of thought and research around the technological applications and biological interactions of metalloids revolves around the six elements commonly recognised as metalloids. As stated the article's primary focus is around the elements commonly recognised as metalloids, simply because of the lack of agreement as to which elements are appropriately called metalloids. Sandbh (talk) 12:58, 22 March 2013 (UTC)

• The discussion of allotropes takes place in a subsection of "Elements less commonly recognised as metalloids" because some allotropes of elements otherwise regarded as either metals or nonmetals are sometimes regarded as metalloids. Phosphorus, for example, is ordinarly regarded as a nonmetal but the black allotrope is sometimes regarded as a metalloid. Arsenic and antimony are included in this section as both elements are sometimes described as metals yet both have allotropes with metalloidal qualities. I understand the point you are raising about arsenic and antimony, however, so I will see if the wording of this part of the article can be better expressed. [Fixed Sandbh (talk) 12:26, 23 March 2013 (UTC)]

4. Notes. The footnotes are an outcome of a long series of judgements designed to achieve an article of reasonable length (earlier reviewers thought the article was too long) combined with useful supplementary material. I agree with you that the sentence in which footnote 38 appears leaves the reader guessing too much ('What do you mean by a single quantitative criterion?') so I'll expand that a bit to include an example. Good call.

Expansion was not the goal I had in mind. In case I wasn't clear, this article has way, way too much text in the notes, including some information that's probably superfluous and some information that probably belongs in the article body. Regardless of where it should be, the net effect is to make the article very hard to read, because in order to follow what's actually going on, the reader needs to hop back and forth between article body and the notes. This is not brilliant prose. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

A casual reader can get the thrust of the article without having to read the footnotes. The footnotes were designed to provide supplementary information for the interested reader. I will check the footnotes again, however, to see if any should be incorporated in the main body of the article. I do not think, for an article of this nature, that there are too many footnotes. I also want to save future readers the pain of having to look up this content again and again in the literature.

5. Formatting of citations.

'I'm not going to quibble with the nonstandard citation format'

• To my knowledge, Wikipedia doesn't have a standard citation format?

There's nonstandard and then nonstandard, but, again, the FA criterion is consistency, so I'm willing to give a pass on the overall formatting. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

'Many sources have publication locations, but some do not (e.g. Aldridge 1998)…'

• Aldridge fixed; my mistake.

'There is sometimes, but not always, a comma before reference numbers (e.g Barfuß et al. 1981)…' [Fixed Sandbh (talk) 12:26, 23 March 2013 (UTC)]

• If you are referring to the superfluous comma at the end of the journal article title for Barfuß, that has now been fixed.

No, I'm saying that sometimes you have a comma before the doi, and sometimes (like on the Barfuß source), there isn't. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC) [Fixed Sandbh (talk) 12:26, 23 March 2013 (UTC)]

'…and some reference numbers are missing (e.g Allen & Ordway 1968)'

• Could you clarify what you mean here? I see nothing wrong with Allen & Ordway. It may be that I cannot see the wood for the trees.

It needs an ISBN. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

Fixed, thank you. Sandbh (talk) 12:58, 22 March 2013 (UTC)

'…or not correctly linked (e.g Askeland, Fulay & Wright 2011).'

• Could you clarify what you mean here? I see nothing wrong with Askeland, Fulay & Wright 2011.

The ISBN is not properly linked to BookSources. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

Fixed, thank you. Sandbh (talk) 12:58, 22 March 2013 (UTC)

'The sort of chained-reference format used for Eberle 1995 and Keller 1985 probably isn't acceptable at the FA level.'

• Given citation bunding is an accepted Wikipedia practice WP:CITECLUTTER I am not sure why this would be so.

This is not what is meant by citation bundling. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

My confusion. I see nothing inconsistent here with Op. cit., given there is only one previously cited work by the authors in question. 12:58, 22 March 2013 (UTC)

'What makes Allenden 2012 a reliable source, and what is its year of publication?'

• Year of publication is 2012, as per 'Allenden 2012'. He is a peer recognised expert silversmith. I include a citation to his blog as he gives a particulary lucid explanation for the increased tarnish resistance of argentium silver. However, he does not appear to have been published. I will delete the footnote for now and see if I can harvest a better source from Argentium sterling silver.

Not sure how I read over the publication year. That said, I still don't think this is a reliable source. Surely, there are scholarly publications or at least reliably published books that address this topic? Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

Allenden deleted Sandbh (talk) 12:26, 23 March 2013 (UTC)

'What makes Brown 2007 a reliable source?'

• A published expert (although he does not clearly list his publications).

Please read WP:SPS, including note 8. I'm not convinced that these self-published sources are by authors that rise to the level of "established experts on the subject matter"; at a bare minimum, they do not appear to have achieved the level of notability to have their own articles. Not everyone who has published papers is an "expert" in the context of that policy. This generally applies to all the sources whose reliability I have questioned. The FA criteria include an expectation of "high-quality sources". These aren't, or at least, certainly do not appear to be. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)

'Gray 2010?'

• A published expert.

'Hoppe 2011?'

• Pretty good German source, but not strictly reliable (as far as I can see) so I will remove and see if I can find a better source. [Hoppe removed Sandbh (talk) 12:26, 23 March 2013 (UTC)]

'Tuthill 2011?'

• A very minor reference, in a footnote, to a mnemomic for remembering which elements are metalloids, as published in the Iolani School Bulletin (est. 1960) by Iolani School, 563 Kamoku Street, Honolulu, Hawaii.

'What is Haller 2006? It's hosted by the Office of Scientific and Technical Information but the reference doesn't cite any actual publication information, and I don't see any.'

• Fixed.

'Why are you using Hawkes 1999 at all, especially over Hawkes 2010, which you also cite (including citing both for the same claim at one point). '

• The two articles include different content and different arguments.

'In fact, several times, there are citations to multiple editions of the same work (e.g Metcalfe, Williams & Castka 1966 / 1974 / 1982); if this isn't necessary, it feels like WP:BOMBARDMENT.'

• This is generally the result of later editions of the same work including new content. Sandbh (talk) 02:16, 17 March 2013 (UTC)

Do later editions of the work also remove existing content? If there's a single edition that has all the material, you should be referencing that, rather than two or three different editions. Squeamish Ossifrage (talk) 18:29, 17 March 2013 (UTC)