Old News

This is a reflection on the technologies we use to read and cite “the news”. For me, that involves both old news and breaking news. In any given week, I’m likely to read this “news” on websites (both open and closed), on newsprint, and on microfilm. I’ll listen to it on the radio and watch it on tv too. Yes, that’s all of the above. For reasons I’ll discuss below, I can’t predict that this complex, multimedia aspect of my reading life will be simplified anytime soon. I’m not advocating any such simplification either, though of course there are things I’d like to see changed.

I doubt I’m a typical news reader, listener, and viewer. Most readers (and many publishers) don’t embrace the oxymoron implicit in the term “old news”. For most, if it’s not new, it simply isn’t news anymore. Once we have been alerted to it, some say, news is disposable. Why would anyone want to archive it, cite it, catalogue it, classify it or index it?

I do understand that point of view. For example, although I acknowledge it does have its uses, I seldom actually do use search.twitter.com (or snapbird.org either). We all have to draw the line somewhere. I still have some tendency, when it comes to “the news”, to draw my line around the “classic” newspaper. I’ll admit that my line is getting fuzzier with each passing day. More about that below.

What I’m looking for, first and foremost, is citability. That means something like accessible permanence. Preferably, that permanence will be the result of many of us wanting to preserve and cite the work of the most honest, persistent and able journalists. There are also, however, topics concerning which we’re happy, whatever the quality of the work, to discover that there was any coverage at all to begin with, and that it has survived until now.

I do know that preservation has a cost. The people with the money for preservation have to draw a line somewhere too. As to accessibility, it depends on such things as the storage media used, and policies with respect to copyright administration. It’s easy enough to predict that there are going to be commercial and institutional “accidents” affecting the permanence and accessibility of much fine journalism.

Paper

When it comes to permanence, something must be said about the paper in “newspaper”. Newsprint isn’t easy to preserve. Much of it has been cheap and disposable. I still pay (out of my own pocket, not institutionally) for print subscriptions to six newspapers. They are all locally owned weeklies. There’s a seventh that I would subscribe to if I could. It’s delivered free within the community where it’s published, but, even though I’d happily pay the postage to get a copy sent to Toronto, putting a copy in the mail every week is not something they’re set up to do. I’m lucky to have a friend who saves the copy delivered to her doorstep for me. Because she occasionally experiences delivery problems, I make the occasional visit to the newspaper office to beg for back issues. Often there are none to be had. Interestingly, this is a paper that didn’t even exist four years ago. At least it has a good website.

Some of these locally-owned newspapers don’t have websites. Others do, but that’s not what they’re about. At this point, I don’t trust some of these papers to archive their digital content (such as it is). A lot of the content from the print editions is never put on the websites in any event.

I wish I had confidence that local and other libraries would collect these papers systematically (in any format), but I don’t. Some local libraries are great, of course, but the quality of local history collections work varies considerably from town to town. I would like to empty my house of some of my personal collections sooner rather than later, but the time doesn’t seem to be right. I’m not at all confident that I will be able to find libraries who will be interested on the day when emptying my house becomes an imperative.

Microfilm

Most local libraries will have at least some of the local papers on microfilm. The Canadian Library Association initiated a newspaper microfilm project in 1947: see Bruce Peel, “The Microfilm Project of the Canadian Library Association” Microfilm & Imaging Review 3(2):106-7 (1974). The Ontario Community Newspaper Association has also, for a number of years, provided a microfilming service for its members, in association with Access-Systems Ltd. Other Canadian companies in the business include Commonwealth Imaging and Preston Microfilming. Of course, they’re all now digitizing images too. ProQuest, the successor of University Microfilms International, bought the Canadian company, Micromedia, in 2002. A number of western Canadian papers have also been filmed or digitized by HeritageMicrofilm and NewspaperArchive, respectively.

The greatest collector of newspapers in Canada is Library and Archives Canada, but they are far from having a comprehensive collection. The guide to their collection, “Newspapers at Library and Archives Canada“, is very helpful. I don’t live in Ottawa, but Toronto isn’t a bad place for newspapers either. The Archives of Ontario, which has a good collection, lists its holdings in Research Guide 212 and in the very large Finding Aid L23. York University has this guide to newspapers, the University of Toronto has that one, and Ryerson another one. None of these guides provides much help with respect to microfilm holdings. The Toronto Reference Library also has a significant collection of newspapers on microfilm.

Digitized Microfilm Page Images

Most of us who dabble in local history have heard the legends about how the old print backfiles were rescued from a dumpster after filming, and stored in the loft of someone’s barn. It’s a lucky community where the legend turns out to be true. I’ve had a number of interesting and informative conversations over the years with Jerry Dupont of LLMC about the reasons for digitizing from print, whenever possible, rather than from microfilm. I can’t remember all his points in detail, but he was persuasive. There is a useful review of the subject in Michael A. Banks, “Small-town newspapers online: smaller newspapers have been slower to jump on the online bandwagon, largely due to the high cost of converting pages to digital format and extracting text with OCR software,” Online 30(1):34 (January 1, 2006).

Some of the problems with digitizing from microfilm are not difficult to illustrate. Sometimes, whether because of the equipment or the operator of it, the microfilming was simply badly done. If the film itself is illegible, then the digitized images taken from it will be illegible too. That means that the quality of optical character recognition will be low, and that the searching will be bad.

ProQuest

Anyone who has used the digitized Globe & Mail (Canada’s Heritage from 1844) or the digitized Toronto Star (Pages from the Past) has seen the problem. Both of these are ProQuest products. Although I suppose as much as could be done with the images available probably was done, these products are sometimes quite frustrating to use.

ProQuest was previously known as Bell and Howell Information and Learning, UMI, and University Microfilms. The larger group of “ProQuest Historical Newspapers” includes the New York Times, the Wall Street Journal and the Guardian.

Gale

In 2007, Thomson Gale was sold to Apax Partners and OMERS Capital Partners: Norman Oder, “Thomson Gale Part of Larger Sale” (libraryjournal.com). The result was Gale, part of Cengage Learning. The Gale NewsVault includes
• Times Digital Archive, 1785-1985,
• 17th and 18th Century Burney Collection Newspapers,
• 19th Century British Library Newspapers Part I and Part II (gale.com), and
• 19th Century U.S. Newspapers.

Readex

Readex, a division of Newsbank, is another microfilm publisher with digital offerings. Its most notable product is Early American Newspapers (newsbank.com), a part of the larger America’s Historical Newspapers (newsbank.com) group of products.

PaperOfRecord and the Google News Archive

One of the particularly interesting stories in this area is that of PaperOfRecord, currently said to be in partnership with Hypernet. (I have no idea what “partnership” means in this context.)

The story began with Bob Huggins, CEO of an Ottawa-based company called Cold North Wind Inc. In July 2000, there was news that the company had plans to create digital images from microfilmed newspapers, in partnership with Bell and Howell Information and Learning: Andrew Flynn, “Read all about it: Web site to archive old newspapers–Ottawa firm gets rights to 11,000 North American papers dating to 1700s,” Toronto Star, July 3, 2000 (LexisNexis). There were plans to be online, under the name “Paper of Record”, by the end of the year. Not long after, it was reported that Torstar had taken a 10 percent share, and that images of the Toronto Star would be digitized: Bob Ferguson, “Torstar invests in Internet firm,” Cambridge Reporter, October 3, 2000 (LexisNexis).

Bell and Howell/UMI had announced a “digital vault initiative” two years before: Paula J. Hane, “UMI Announces Digital Vault Initiative,” July 13, 1998 (infotoday.com). In 2001, Bell and Howell Information and Learning became ProQuest Information and Learning: Paula J. Hane, “Bell & Howell Becomes ProQuest Company,” June 11, 2001 (infotoday.com).

By the spring of 2003, ProQuest was described as a “former investor”. Cold North Wind had purchased the rights to 250 defunct newspapers, and PaperofRecord.com, launched in November 2002, had quickly gained 10,000 subscribers: Bruce Gillespie, “All the news that’s fit to scan,” National Post, April 7, 2003 (Westlaw). The Toronto Star project had been completed, and TorStar had a 19 percent stake: Bruce Gillespie, “History pays,” National Post, December 1, 2002 (Westlaw).

At that point it sounded like a happy story. Of course, there were copyright issues, for which see Freelance Writers Class Action (Koskie Minsky LLP). But a good revenue stream should ultimately help to solve most copyright problems.

Unfortunately, the financial challenges PaperofRecord.com was facing were revealed in late 2008, when it completed a deal with Google for 20 million newspaper pages. A deal to sell the database to Library and Archives Canada had fallen through in late 2003. Neither a paid subscription model nor a fee for advertising model had been successful. From 2005, PaperOfRecord had been supported as a free site under a financing agreement with Google, even as it continued to digitize new material under an exclusive licence from the Canadian Library Association: Bert Hill, “Google expected to take over Ottawa data firm,” Ottawa Citizen, November 14, 2008 (Westlaw).

Google’s new initiative was the Google News Archive, described in “Bringing History Online” (September 2008) googleblog.blogspot.com. The titles acquired were listed here: “Paper of Record titles acquired by Google” (google.com)

Unfortunately, the transition was not a smooth one for PaperOfRecord users, as Robert B. Townsend of the American Historical Association reported: “‘Paper of Record’ Disappears, Leaving Historians in the Lurch,” (May 2009) (historians.org). According to Townsend:

In light of the complaints of historians and others, Bob Huggins, the founder of Paper of Record, received permission to re-start the service. According to Huggins, the revived resource will include all of the content that was taken down, but it will only be available through subscriptions by academic institutions.

In May 2011, Google stopped building the news archive. AFP reported, in “Google stops digitizing old newspapers,” May 20, 2011 (google.com):

People will still be able to find newspapers already converted to digital format in the Google News Archives at news.google.com/archivesearch but the collection won’t grow. …

Digital copies of archived newspapers will be given to respective publishers that will be free to host the issues on their own websites.

According to Carly Carioli, “Google abandons master-plan to archive the world’s newspapers,” May 19 2011 (thephoenix.com):

The deal Google struck with partner newspapers stipulated that, somewhere down the line, a paper could purchase Google’s digital scans of its content for a fee. That fee is now being waived, and Google is not only giving publishers free access to the scanned files, but also the rights to publish them with other partners.

She speculated:

Google reportedly used its Maps technology to decipher the scrawl of ancient newsprint and microfilm; but newspapers are infamously more difficult to index than books, thanks to layout complexities such as columns and jumps, which require humans or intense algorithmic juju to decode. Here’s two wild guesses: the process may have turned out to be harder than Google anticipated. Or it may have turned out that the resulting pages drew far fewer eyeballs than anyone expected.

Bob Huggins’s reaction to this news was reported by Vito Pilieci, “Google archive decision ‘astonishing’: End of newspaper digitization effort disappoints originator of technology,” Ottawa Citizen, May 25, 2011 (Westlaw)(ottawa.citizen.com)

The Google titles are listed and available for browsing, at least for now, at http://news.google.com/newspapers.The fee schedule for access to PaperOfRecord on the hypernet.ca website is at https://paperofrecord.hypernet.ca/register.asp.

Some Other Initiatives

Other interesting newspaper initiatives include the digitized Journaux to be found in the collection numérique of the Bibliotheque et Archives nationales Quebec. For all the lawyers, I note especially the Gazette officielle du Québec 1869-1995. A similar project is Early Alberta Newspapers (ourfutureourpast.ca). Again, for the lawyers, I note the digitization of the Alberta Gazette. Library and Archives Canada has digitized the Canada Gazette.

Still growing is ink.ourontario.ca, a project of Knowledge Ontario. If I’m not mistaken, a lot of the initiative for this came from the Halinet, a group of Halton historical and genealogical societies.

In the United States, the Library of Congress supports an initiative known as Chronicling America.

Born-digital Page Images

Prints of page images can be purchased from the Globe & Mail ArtStore. For prints from page images prior to 2003, orders are filled by Preston Microfilming. From 2003 on, the prints are said to be in full colour from high quality PDFs. So something significant happened in 2003.

First, then, a bit of historical perspective. Once upon a time, there was a device known as the printing press.

We tend to associate the printing press with Gutenberg. The process of printing, however, is thousands of years older than Gutenberg. Woodblock printing seems to have been the earliest method. Two techniques later developed for creating images on metal for use in printmaking were engraving (with a metal burin or graver) and etching (with wax and acid.)

Gutenberg’s innovation, better described as letterpress, involved the use of moveable-type in typesetting, and the use of the screw press in printing. The creation of images with which to illustrate a text continued to be quite labour intensive until the invention of photoengraving in the mid-19th century. This was a chemical process which involved the application to a metal surface of light-sensitive materials which, when developed, would resist an acidic etching of the underlying metal. The use of a screen made dotted halftone images possible.

About a century and a quarter ago, hot metal typesetting had a big impact on the newspaper industry. Both letterpress and hot metal are relief printing technologies, in which paper receives ink from the raised parts of a metal surface. Another major method is intaglio printing (or gravure), in which the paper receives ink from the lowered parts of a metal plate. (The corresponding photographic process is photogravure.) A third major method, planographic printing, involves a flat surface. For example, classic lithography, as developed just over 200 years ago, involved the etching of a wax or oily coating applied to a flat limestone surface, instead of the etching of a metal surface.

In the early 20th century, most newspapers switched from relief printing to offset printing. This is a process for reproducing master images lithographically. (The master images might have been created in any number of ways.) There are methods of chromolithography for colour printing, but this was too expensive to be used with disposable newsprint until relatively recently.

Hot type yielded to cold type with the advent of phototypesetting. According to Wikipedia, “[t]he use of phototypesetting grew rapidly in the 1960s when software was developed to convert marked up copy, usually typed on paper tape, to the codes that controlled the phototypesetters.” Desktop publishing followed in the 1980s, and grew in sophistication. Adobe was a desktop publishing leader, producing the PostScript page description language and the Personal Document Format (PDF). The W3C has even entered the field with XSL Formatting Objects.

And so, finally, all this means that most page images are currently born digital. There shouldn’t be any need to digitize from print originals. The interesting question then–and I don’t pretend to have all the answers–is why we don’t see more retrospective collections of these digital page images on the web.

Virtual Paper/Papier Virtuel

I can only speak to my personal experiences, which are unfortunately very limited. My university, as far as I know, doesn’t provide access to any of these systems, so I have to pay for them myself. I subscribe to the electronic-only edition of The Hamilton Spectator, and I subscribe to PressDisplay from NewspaperDirect.

The Hamilton Spectator is part of the Metroland Media Group owned by Torstar Corporation. The Record (Waterloo Region) and the Guelph Mercury are part of the same family, and appear to have similar editions. Oddly, I couldn’t find a distinct, electronic-only edition of their cousin, the Toronto Star. (As indicated below, Toronto Star page images are available as part of PressDisplay. However, the page images of its Metroland cousins are not available on PressDisplay.)

I don’t have a record of when I first started routinely reading Hamilton Spectator page images online, but at that point pdf images were available for free. The free part changed in early 2009. At the same time, because of the peculiar system chosen to keep the non-subscribers out, Internet Explorer became the only browser that I could use to access the pdf images. Moreover, even with a password and IE, I could only access the site from one of the other of only two different ip addresses. What a pain. In the summer of 2010, however, life became much better. The company switched from what looked like a home-grown digital publishing system to VirtualPaper, also known as PapierVirtuel.

VirtualPaper, which is based in Montreal, claims to be “the #1 digital publishing solution available today.” They say “Our publications are displayed in a user-friendly reader interface which uses either standard HTML or Adobe Flash.” Also, “Get your document on the iPad and other portable devices, without any app or plug-in to install. That’s the virtual Paper way! We use HTML5 technology to display your content which means practically all tablet and portable devices can access your publications instantly.”

NewspaperDirect/SmartEdition/PressDisplay

If VirtualPaper is “the #1 solution” (which I imagine must be debatable), a big part of the problem nowadays is that people need Kindle newspapers from Amazon, Kobo newspapers from Google, Nook newspapers from Barnes & Noble, and Blackberry newspapers from Blackberry, as well as IPad newspapers from Apple.

There are other companies in the business. These include the previously noted NewspaperDirect of Vancouver BC, NewsStand of Austin TX and OliveSoftware of Denver CO.

I can’t say much about things I haven’t used, so I hope others will pitch in. There are probably other companies involved in this business too that I haven’t found.

NewspaperDirect’s publishing service to newspapers is called SmartEdition. It is used for Globe2Go, the Globe & Mail-branded version of that paper; for the National Post and other papers in the CanWest family, branded as DigitalNationalPost; for the Toronto Sun-branded eedition of that paper; and for the version of the Toronto Star available on PressDisplay, noted above. (There will be a general description of PressDisplay below.)

NewspaperDirect’s PressReader software can be used for offline reading. PressReader apparently runs on Android, Honeycomb, Blackberry, IPad/IPhone, webOS and Windows. (See the website and blog).

I’ve noted that Apple, Amazon, Google, Barnes & Noble and Blackberry are selling newspapers directly to readers under their own brands. NewspaperDirect (under the name PressDisplay) also has a large international repertoire which includes many Canadian papers. NewsStand sells directly to readers too, but has just a few Canadian papers: only the Globe & Mail in English, and several in French. All of the Canadian papers on NewsStand are also available on PressDisplay.

My Hamilton Spectator subscription costs $5.00 per month. The Globe & Mail would cost $159.50 per year prepaid (not quite $16.50 per month) on Globe2Go (via NewspaperDirect). On NewsStand, it would cost $20.00 CAD per month.

The pricing on PressDisplay is more complicated. A personal subscription is $29.95 USD per month, a corporate subscription $99.95 and a professional subscription $199.95. As to what is included, you need to read the fine print. For example:

Excessive or Unreasonable Use
(a) In the case of usage of the Service, excessive use will be use above 100 issues of publications under Personal account per User per month, above 250 issues of publications under Corporate account per User per month, above 500 issues of publications under Professional account per User per month.

My routine use of PressDisplay, at one point, was to read five dailies and one weekly. When the “unreasonableness” of this was pointed out to me, I opted to cut back and read just the papers that were only available on PressDisplay. Interestingly, it would cost less to have three personal subscriptions than to have one corporate subscription.

You can do monitoring with alerts with PressDisplay. The number of stored queries you are allowed also depends on the type of subscription.

The number one problem with both my NewspaperDirect and VirtualPaper subscriptions is that the papers are only accessible for a couple of weeks after publication. What that means, of course, it that these services are essentially uncitable as such. The only sensible thing to do is to forget about linking, and cite to paper or microfilm. Once again, the working assumption seems to be that news is disposable. (The papers would be accessible for a couple of months in the case of a corporate or professional subscription to PressDisplay, but that makes no practical difference to me.)

Article-oriented Websites

The Classics

As a member of the York/Osgoode community, I’m lucky not to have to pay for my academic use of the news articles on LexisNexis, Westlaw, ProQuest and Factiva. All these services provide access to databases of full-text newspaper articles. There’s boolean searching, and you can set up monitoring with alerts. (Another service called Eureka.cc is sometimes available to me too, but not currently. It has some PDF content in addition to its text content.)

These services had their origins in the 1970s, and were a natural byproduct of digital typesetting processes. Indeed, that’s why some of the databases have articles from that era, which is great.

You’re not likely to find anything but text in these databases, which is too bad. The photographs and illustrations might be described, but they won’t be available for viewing.

Typically, only the articles are included. Not the advertising or the classifieds. That’s a shame too.

The content is marketed in various ways. Have a look, for example, at this little bit of cooperation between the Toronto Star and ProQuest.

It’s possible to create hypertext links to the articles on these systems for my personal use, but none of the systems is entirely satisfactory when it comes to making hyperlinks which will work for broader audiences. ProQuest is probably the best in this respect, but there are the usual problems making a link work for ProQuest customers who need to access the system through a proxy server. Westlaw and LexisNexis come close too; the biggest problem is the national customization of the interfaces: for example, a hyperlink which works fine for other Canadian customers might be completely useless to a U.S. customer. Factiva has done something very odd too by requiring that there be a customer-specific authentication string in hyperlinks.

Individual Newspaper Websites

Every business needs a website these days, and newspapers are no exception. Some of them have great websites. Some even use the services of companies like NewspaperDirect and VirtualPaper for article-oriented websites as well for page-image websites.

My impression is that most newspapers still consider the versions of the articles in their printed papers and page-image websites to be more definitive than the versions on their non-page-image websites. More and more, however, it appears to be a case of several products developed in parallel that just happen to have a lot of content in common. It would be interesting to have the perspective of the newspapers’ management accountants as to how the different products perform.

Some of the big differences? The look and feel of the medium, of course. Timeliness of breaking news. Hyperlinks to related content. A multitude of reader comments versus a small selection of letters to the editor. Audio and video content. I’m sure there are things I’ve forgotten. I’ll have something to say about citability below.

Radio and Television News Websites

I referred to audio and video content in the last paragraph. It is worth remembering that there has been radio and television coverage of the news for many years. Until the web came along, however, newspaper news was a lot easier to research than radio or tv news. The good news is that organizations like the CBC do have a lot of news text on their websites. The best way to look for news articles on the CBC website, I think, is with Google and +site:cbc.ca. If searches for Gretsky and Chretien are indicative, there is web-content from January 1998 forward. I have no idea what relationship, if any, there is between these articles and anything that’s actually broadcast. It certainly isn’t made obvious.

As far as I know, searching for audio and video news broadcasts is still mostly a matter of searching for textual descriptions of them. That, at any rate, is the kind of searching available at archives.cbc.ca and archivesales.cbc.ca. Yes, there are cool things out there like TinEye, PicTriev and Midomi. The people working with archival collections probably have much better tools. I’ll just admit it’s an area with which I’m not very familiar.

There doesn’t really seem to be a lot of material on archives.cbc.ca. There is much more to be found on archivesales.cbc.ca, a service for commercial producers only. As indicated on the CBC Transcripts and Tapes page, ordinary folks are asked to deal with Cision, a public relations firm.

Westlaw has transcripts of the television and radio news broadcasts of a number of CBC, CTV and Global stations and cable channels. The CBCCBLTTVTCA database, for example, has transcripts from CBLT TV in Toronto. The CBCRAD1TO-CA database has transcripts from CBC Radio 1 — Toronto. ProQuest’s CBCA database has some transcripts too. I suspect LexisNexis has more than is visible to the eyes of a mere academic.

Citing the News

I’m not saying it’s perfect, but I have a standard practice when taking notes from news sources. If I’m looking at the printed page of a newspaper, or at a page image on a website, I note the date of publication, the name of the newspaper, the page and column, the headline, and the byline, at least. I also construct links to LexisNexis, Westlaw and ProQuest when I know there are corresponding databases. I can usually construct those links without actually signing on to the systems. I can’t construct a link to Factiva without being signed on to it, so I only construct links to Factiva selectively. Another thing I only do selectively is to look for and link to the same story on the newspaper’s own website.

My routine is slightly different if I have begun by looking at the newspaper’s website. I make a link, of course. I note the name of the newspaper, the headline and the byline. I guess at the date when the article might appear, or might have appeared, in the printed newspaper. Of course, there are no pages or columns to note. I do try to construct links to LexisNexis, Westlaw and ProQuest. Selectively, I’ll follow one or more of those links to get a page number, and to check the date of publication, the headline and the byline. Selectively, as well, I will sign on to Factiva to construct a link, or sign on to a page-image database to check out the photographs or illustrations, or to note pages and columns.

There’s a lot of overhead built into these routines, but maybe less than you imagine. Constant repetition seems to make me speedier.

When I review my notes from time to time, it’s pretty handy to have a selection of links to choose from. Newspapers do have an annoying tendency to come and go where particular aggregators are concerned.

The articles have an annoying tendency not to be the same on the website and in print. The headlines seem to be different more often than not, the datelines can differ by as much as several days, and it’s surprising how often even the bylines are not the same.

As noted above, my impression is that what appears on the website tends to be the first draft, and that the publishers still regard the printed page as more definitive. In principle, however, it makes more sense to think of the website and the printed newspaper as distinct publications that just happen to have a lot of common content. For sure, people should not just assume they are the same.

How Things Have Changed

Just for fun, I thought I would see how The Chicago manual of style has changed over the years when it comes to citing newspapers.

The 16th edition of The Chicago manual of style (Chicago: University of Chicago Press, 2010), para. 14.203, says:

The name of the author (if known) and the headline or column heading in a daily newspaper are cited much like the corresponding elements in magazines (see 14.199-202). The month (often abbreviated), day, and year are the indispensable elements. Because a newspaper’s issue of any given day may include several editions, and items may be moved or eliminated in various editions, page numbers may usually be omitted (for an example of a page number in a citation, see 14.209). In a note or bibliographical entry it may be useful to add "final edition," "Midwest edition," or some such identifier. If the paper is published in several sections, the section number or name may be given (e.g., sec. 1). To cite an article consulted online, include the URL; in some cases, it may be advisable to shorten a particularly unwieldy URL, to end after the first single forward slash (i.e., the slash that follows a domain extension such as .com).

The example given in 14.209 is “45. David Frum, “The Vanishing Republican Voter,” New York Times Magazine, September 7, 2008, New York edition, MM48.”

The 15th edition of The Chicago manual of style (Chicago: University of Chicago Press, 2003), para. 17.188, says:

The name of the author (if known) and the headline or column heading in a daily newspaper are cited much like the corresponding elements in magazines (see 17.183-86). The month (often abbreviated), day, and year are the indispensable elements. Because a newspaper’s issue of any given day may include several editions, and items may be moved or eliminated in various editions, page numbers are best omitted. In a note or bibliographical entry it may be useful to add "final edition," "Midwest edition," or some such identifier. If the paper is published in several sections, the section number or name may be given (see examples in 17.191). For papers published on the Internet, adding a URL will show that an online edition was consulted (see 17.198).

The 14th edition of The Chicago manual of style (Chicago: University of Chicago Press, 1993), para. 15.234, said:

In citations to items in daily newspapers, the day, month (often abbreviated), and year are essential. Because items may be moved or eliminated in various editions, page numbers are usually omitted. For a news item in a large city paper that prints several editions a day, the name of the edition is useful (first edition, city edition, late edition, etc.) because the item may not appear in all editions. (Do not confuse edition with issue, which means any edition published on a specific day. Thus, a newspaper’s issue of 22 February 1980 might consist of several editions.) The name of the edition is not included in references to editorials, features or other material that appears in all editions of the day. If page and column numbers are included, use p. and col. to avoid ambiguity–for example, p. 3, col. 4. The citation should also include the author’s name and title of the article, if these are given.

The 13th edition of The Chicago manual of style (Chicago: University of Chicago Press, 1982), paras. 17.57-58, said:

In references to daily newspapers the day, month (usually abbreviated), and year are essential; page numbers are usually omitted. For a news item in a large city paper that prints several editions a day, the name of the edition is useful (first edition, city edition, late edition, etc.) because the item might not appear in all editions. (Do not confuse edition with issue, which means any edition published on a specific day. Thus, a newspaper’s issue of 22 February 1980 might consist of several editions.) The name of the edition is not included in references to editorials or features or other material that appears in all editions of the day. If page and column numbers are included, use p. and col. to avoid ambiguity–for example, p. 3, col. 4. …

References to papers published in sections–almost all Sunday newspapers and large daily newspapers such as the New York Times and Chicago Tribune–usually include the name or number of the section. …

The 12th edition of The Chicago manual of style (Chicago: University of Chicago Press, 1969), para. 15.123, said:

References to daily newspapers or weekly publications require only the date of issue–day, month, and year. Volume numbers are unnecessary. In newspaper references it is best to omit page numbers also; in different editions of the same issue a particular news item may not be on the same page in each edition. It is useful, however, to give page numbers for references to the New York Times and the Times (London), citing the edition consistently used in preparing the indexes to these two newspapers.

The 11th edition of The Chicago manual of style (Chicago: University of Chicago Press, 1949), p. 149, said:

Refer to a newspaper by the title and date. If the title does not include the place of publication, the name of the place should be given in parentheses after the title. Do not include the definite article as part of the title. Page numbers are recommended: … Times (London), May 19, 1946, p. 4, col. 2.

In some editions, it has seemed more important to insist upon the complexity of newspaper publishing practices. In other editions, there appears to have a been a desire to lighten the burden of writers and editors. Interesting.

I haven’t reproduced the sections on the generic citing of websites. My only criticism of the 16th edition is that it might have highlighted, for the benefit of less careful readers, the fact that there may be several quite different referents for the phrase “an article consulted online”. As noted above, an article on the newspaper website should not uncritically be cited as if there were bound to be an identical article in the newspaper.

Conclusion

Sadly, although the article that appears on the page-image website is typically the best-edited version of it, complete with illustrations and photographs, the links to it will typically only persist for a couple of weeks, making that version practically uncitable. And although the aggregator’s version of the article, despite lacking photographs and illustrations, is typically a better-edited version of it, the less well-edited version that appears on the newspaper’s own website is typically the one that’s best to link to and cite when writing for a broad audience. Doesn’t anybody think about these things? Or is there a great big part of the puzzle that I haven’t yet stumbled across?