Talk:Language/GA1

GA Review

Article (edit | visual edit | history) · Article talk (edit | history) · Watch

Reviewer: Chiswick Chap (talk · contribs) 13:04, 6 September 2014 (UTC)

I'll take this on. Quite a topic. Chiswick Chap (talk) 13:04, 6 September 2014 (UTC)

General comments:

"...many languages...", "several languages...": Wikipedians are wary of such generalisations. I hate language lawyering (ahem, given the article's subject), but these phrasings are probably best avoided, using such things as "some languages..." unless you're really sure in a specific case (when "most languages" or "Indo-European languages" or whatever would be better).

"today", "currently": these quickly go / may have already gone out of date. Please replace such time-dependent words or phrases with specific dates.

Specific comments:

There needs to be a sentence or two on the philosophy of language in the lead, to summarize (well, at least to mention) the range of views discussed in the Definitions section.

Definitions, Main article: Philosophy of language: I think this section is probably underweight, given the range and strength of views on the subject. For FA the article will definitely need to cover more of the philosophy of language; for GA, Kant/Chomsky/Fodor are probably covered sufficiently, as are de Saussure and Wittgenstein. W.V.O. Quine does I think deserve a mention, however; and it might be best at least to mention the theories of empiricism and of Tarski briefly.

Unless Maunus beats me to it, I'll try to improve this at the library this weekend. Not my strongest suit, though.

Peter ^Isotalo 18:46, 18 September 2014 (UTC)

I've done some work on that. See if you like it.User:Maunus ·ʍaunus·snunɐw· 23:38, 18 September 2014 (UTC)

Definitely an improvement. Is Tarski going to get a line or two? Chiswick Chap (talk) 09:19, 19 September 2014 (UTC)

I think Tarski could only be briefly mentioned in the section on language as a formal symbolic system. I've added a sentence on formal logic.User:Maunus ·ʍaunus·snunɐw· 14:08, 19 September 2014 (UTC)

OK. Of course that might need a ref too. Chiswick Chap (talk) 14:20, 19 September 2014 (UTC)

Mental faculty: do we need to have the word "often" in there (twice)? Seems a bit weaselly. By the way, can we do better than "One definition sees..." - could we not say something like "The mental faculty view of language ..."?

Mental faculty: "Kant and Descartes": perhaps say "the philosophers...".

The lead image in the Syntax section correctly describes predicate (grammar) in the caption but fails to show it in the image. It would be best if the label "Predicate" could be added to the image at the (unlabelled) node above "Prepositional phrase". This would both clarify the structure of the tree shown in the image and match the caption.

Anatomy of speech: there seem to be too many uses of italics here. Scientific English words like uvula should not, I think, be in italics (they're just English words, as confirmed by the lack of italics for that article's title). Please review all the uses of italics in the section (if not the whole article) and remove all that are not essential.

"The study of the genetic bases for human language is still on a fairly basic level": suggest link to genetics (or something more specific), and avoid the word "basic" (and probably "fairly", which sounds OR-ish) in the context. "positively implied" -> "definitely implicated".

"fossils can be inspected to look for traces of physical adaptation to language use". Some examples of what such traces might be would be worth giving (not obvious); in fact, this would be a good place for a photograph of such evidence.

" Often, semantic concepts are embedded in the morphology or syntax of the language in the form of grammatical categories". Please provide a brief example (could with benefit be an image with caption). Actually I'm unsure what is meant by categories here. Pinker in Words and Rules p4 says "the word's part of speech, or grammatical category, which for rose is noun (N)", whereas the WP article says "Categories may be marked on words by means of inflection." Which is meant here?

I think this is just Pinker using the word "grammatical category" in a non-standard way using it as synonymous with syntactic category (which perhaps some people do in the generative tradition). In basic linguistic theory grammatical categories are all semantic and syntactic categories that are morphologically or syntactically marked. I will try to think of an example.User:Maunus ·ʍaunus·snunɐw· 15:19, 12 September 2014 (UTC)

"the beginning of the Bronze Age in the late Neolithic period of the late 4th millennium BC": there's something wrong here. Suggest remove mention of Neolithic unless a specific point is being made about overlap of ages, in which case the word "overlap" should appear somewhere.

"Language change may be motivated by "language internal" factors, such as changes in pronunciation motivated by certain sounds being difficult to distinguish aurally or to produce, or because of certain patterns of change that cause certain rare types of constructions to drift towards more common types" uses "certain" three times. Please rephrase.

Language contact: does this section need four "main" links? Language contact already links to pidgin, creole etc.

BTW the formatting of names in citations is not a GA matter, but "de Saussure" is certainly the man's surname. He wasn't christened "De". Chiswick Chap (talk) 15:16, 18 September 2014 (UTC)

That is correct, but I think he is usually cited as Saussure, not de Saussure. And he is always alphabetized under S.User:Maunus ·ʍaunus·snunɐw· 16:15, 18 September 2014 (UTC)

Just as we list books and plays that begin with The or A under the next word. But that still doesn't make "de" a forename. Chiswick Chap (talk) 16:34, 18 September 2014 (UTC)

Just as putting it within the field "first=" in the reference template also doesnt make it a first name. It is just the easiest way to make the template syntax work while citing him as "Saussure" instead of "de Saussure". User:Maunus ·ʍaunus·snunɐw· 16:56, 18 September 2014 (UTC)

Oh dear, what a hideous kludge. Still, worse things happen in databases every day. Chiswick Chap (talk) 17:20, 18 September 2014 (UTC)

Images:

Images should generally be of default width, i.e. should use "thumb" with "upright" as needed; this has the clear advantage of resizing automatically for different user preferences, providing a measure of future-proofing also. I understand the impulse to make maps in particular appear large, but their thumbnails cannot be made large enough to be fully readable, and giving a fixed size is awkward for users of mobile devices, and fails to accommodate change and user preference.

The de Saussure image is out of copyright in Europe but its USA status is unclear, so this needs fixing (on Commons).

The descriptions of the Hangul wi symbol and the KSL wi hand gesture as spectrograms on Commons are wrong. (this is a display bug)

All other images ok (all on Commons).

References:

The citations are in 'Author (year:page range)' format, with 7 exceptions. This does not matter for GA but would need to be tidied up for FA.

I have marked up places where citations are needed in the main text.

Quite a number of references (e.g. 10 Saussure, 13 Chomsky, 21 Tomasello, 22 Deacon, 34 Trask, 72 Bauer, 73 Haspelmath, 80 Nichols, 80 Comrie, 82 Greenberg, 84 Campbell, 91 Kennison, 93 Foley, 94 Agha, 97 Aitchison, 102 Labov, 103 Labov, 108 Thomason & Kaufman, 108 Thomason, 109 Matras & Bakker, 110 Lewis, 113 Katzner, 113 Comrie, 113 Brown, 114 Moseley) are to entire books. This may be appropriate in some cases (e.g. to the general drift of The Language Instinct) but looks wrong in other cases, where suitably narrow page ranges are needed. Especially in cases like the multiple uses of 113 Katzner, Comrie and Brown, separate page ranges are probably required for each usage of each book, rather than lumping the whole of three entire books together. This could be a critical issue for GA.

On reflection I think most of these are not critical for GA, but will need fixing for FA. That said, properly filled-in page ranges are highly desirable, and many editors expect to see them throughout. Chiswick Chap (talk) 06:50, 12 September 2014 (UTC)

Actually most of these are meant as sites to the entire book, as a source for more information rather than as specific cites for a piece of information. For example when I wrote the sentence on "mixed languages" giving Michif and Kriyol Ayisiyen as examples. I referred the reader to the three main works on contact linguistics Kaufman and Thomason, Matras and Bakker, and Thomason. The reason I do this is 1. to lead the reader to the best available literatiure on the topic, 2. because each sentence is a highly condensed description of an entire field of linguistics, meaning that it contains information that is not found on a specific page of the cited works but which is a summary of the cited works. Similarly the references to Chomsky and Labov are references to entire works, because the statement they support is meant as a summary of those works. I will look through the references to see if some of them can be replaced by narrow ranges or page numbers, but I think that in writing I have supplied pagenumbers for all the references that I mean to cite as specific pages, and not done so whenever I meant to refer the reader to broader literature. I know some FA referees are anal about page numbers and dont understand that it is often appropriate to refer to entire works, but oddly I dont find this problem with peer reviewers in journals, who tend to have no problem with referring to full works when not citing a specific fact but pointing towards a set of ideas (e.g. "Evolutionary theory is the hypothesis all known life forms share a single common origin (Darwin 1858)"). User:Maunus ·ʍaunus·snunɐw· 15:07, 18 September 2014 (UTC)

I'm pretty much with Maunus on this one. A quality article about such a general topic is a rarity. GANs, and especially FACs, deal with articles are built up of facts and statements that are seldom, if ever, the topic of entire books. This article, however, is full of very general statements that are very hard to reduce to page ranges. Any reader who will actually want to verify what is being said will easily be able to do this with the reference provided. This is especially true when it comes to sections like "Language families of the world". Works like Katzner (2009) or Comrie (2009) are basically miniature encyclopedias with separate entries for each language or language family. Anyone who can find that book will be able to find what they need even without specific page refs.

I'm going to spend some time at Stockholm University Library this weekend. I'll go through the general refs and see if anything can be improved overall. Right now, however, I can only really agree that (from the list above) Thomason & Kaufman, Thomason, Matras & Bakker and Lewis could use some specifications. Chiswick, if you disagree or believe there I'm missing specific statements lacking specific refs, lemme know and I'll see what I can dig up. The SU library isn't half bad.

Peter ^Isotalo 18:40, 18 September 2014 (UTC)

I basically agree with Maunus also. As I wrote above, I think the 'general' reffing is fine in some cases, more doubtful in others, and while desirable, not critical here at GA. However GA reviewers always raise the matter in my experience, as do ordinary editors when one supplies a wide page range, and rightly so. I would like to direct your energies to closing out the remaining items, leaving the references for FA. I will pass the article here as soon as the remaining non-reference items are closed out. Chiswick Chap (talk) 18:46, 18 September 2014 (UTC)

The Newmeyer 2005 dead link would be best fixed using Wayback Machine or a similar archive service, or else replaced. Needed for FA.

I've added the live link, it was just moved on LSAs webpage.User:Maunus ·ʍaunus·snunɐw· 15:07, 18 September 2014 (UTC)

Overall:

This is an elegantly written and well-structured article. A few issues remain to be fixed. The work of bringing major (top-level) topics to GA is very welcome. Chiswick Chap (talk) 13:28, 6 September 2014 (UTC) I am now happy to pass this article at GA. Thank you to everyone who contributed. Chiswick Chap (talk) 14:34, 19 September 2014 (UTC)

Discussion

Thanks for taking the time to review. A few comments to begin with:

• "Many/several languages": statements in connection with these terms are so general that it might be difficult to specify them. The point of these statements is to show various aspects of how languages can work, not to make claims about how common they are. I think it's better if you try to identify specific problems, because all of them can't be rendered more specific.

I identified two specific ones which I'd like you to change by removing the weasel-like terms; I suggest you also scan the article for other possible problems.

• "Today/currently": pretty much the same as above. Statements like "Linguists currently recognize many hundreds of language families" is not something that is helped by an "as of 2012"-type specification. It will most likely never change. Which ones can actually go out of date in your view?

Then simply remove the "currently" please, rather than arguing. I've taken out some of them now. Chiswick Chap (talk) 12:59, 7 September 2014 (UTC)

• It was published before 1923, so I believe it's PD in the US as well (added additional license).

OK.

• I don't follow your comment about Korean wi. What's wrong on Commons and why is it relevant to this article?

Ah, it's a bug in the multiple-image display, not on Commons, which displays the first caption as if it were a description of the later images when one of these is selected and displayed. The only thing we could do about it would be to avoid using the multiple-image mechanism. I can't say I like it but its use is not inappropriate here.

Peter ^Isotalo 11:49, 7 September 2014 (UTC)

I believe these 'Discussion' items are all satisfactorily resolved in the article now. Chiswick Chap (talk) 13:06, 10 September 2014 (UTC)

Thank you for striking out stuff. I was having some difficulty keeping track of stuff. I'm currently looking into your comments about general refs. To my great chagrin, Comrie (2009) was nowhere to be found at the Stockholm University Library. :-(

Peter ^Isotalo 18:21, 10 September 2014 (UTC)

@User:Maunus, User:Peter Isotalo: this GA process is in danger of timing out for lack of activity. We need to round this off now; if you need more time I can put it on hold. Please let me know your timetable for closing out the remaining issues. I would also say that while the issue of page ranges is not necessarily a show-stopper, I am surprised that no action has been taken to supply any of them at all; most GA articles are fully cited with exact page numbers. Chiswick Chap (talk) 08:02, 18 September 2014 (UTC)

I am not really a part of the review process except in a supportive function. I am a way from my books and cannot help with the page ranges.User:Maunus ·ʍaunus·snunɐw· 14:45, 18 September 2014 (UTC)

The support is valued. Chiswick Chap (talk) 14:58, 18 September 2014 (UTC)

I second that.

Peter ^Isotalo 18:41, 18 September 2014 (UTC)

Sorry for dawdling. I had actually gone through list of page range issues, but got the impression that you didn't deem it necessary for GA status. I'll make a full reply regarding this and the other outstanding issues later today.

Peter ^Isotalo 08:45, 18 September 2014 (UTC)

Lead sentence

The first sentence of the lead has been changed, mid-review, from "Language is the human capacity for acquiring and using complex systems of communication, and a language is any specific example of such a system." to suggesting that language is a "tendency". Since "human capacity" seems to this reviewer to be exactly correct, and to reflect the body of the article, I'd suggest we put it back as it was, unless there are strong reasons to the contrary. Chiswick Chap (talk) 13:00, 10 September 2014 (UTC)

(discussion below copied from user page)

Human beings have the capacity to acquire written language. However, anybody defines written language as "the human capacity for acquiring writing systems". All the best, James343e. 10 September 2014 (UTC)

OK, since we agree on 'capacity', I assume we can revert the text to that effect. Thanks. Chiswick Chap (talk) 14:13, 10 September 2014 (UTC)

"Tendency" is not an accurate way of describing what is basically universal to all humans. "Capacity" is clearly the most accurate summary of the content of the article (and the general opinion of linguists).

Peter ^Isotalo 18:23, 10 September 2014 (UTC)

• The word capacity is used as a theory neutral way of referring to what Chomsky calls the Universal Grammar and the Language Acquisition Device, and which Pinker calls the language instinct and others call. Many other linguists would disagree with this formulation but would agree that there is a certain innate capacity for language that is prior to the acquisition of any specific language. This can also be called the language faculty, but this phrasing I think would be more confusing for the lay reader. That is why capacity is both necessary and a better formulation than for example "tendency" or "ability".