Witty’s Blog

If you are a GLAM looking to make your photographic collection more widely available online, for the last couple of years your first choice would have been to head over to “Flickr Commons”. And you would be in good company too.

However, at least for the current year, Flickr Commons is officially full:

Following a flurry of tweets - led by Mia Ridge who put out a blogpost on this topic much faster than me :-)  - May I take this opportunity then to extend an offer to all of those in “the current backlog” that Wikimedia Commons is open for business - and with a couple of new tricks up our sleeve too.

1) Disk space on the image servers has been dramatically increased very recently. It was getting pretty close to the limit for a while and some MAJOR content donations had to be put on hold whilst that was sorted out. They’ll be announced shortly and I’m really looking forward to it (hint: it’s those Dutch again!) I can’t think of a pretty picture to illustrate this point so I’ll point you to the page that wins my personal “the thing that is quite clearly important but I’m not really sure what it means, award” - http://ganglia.wikimedia.org/

2) The Multimedia Usability project is coming along nicely. Whilst I must admit the Wikimedia upload interface is not as shiny and friendly as the Flickr one, we’re doing our level best to make it easier and cleaner. One of the bigger headaches in improving Wikimedia Commons uploading is that Wikimedia only allows “free content” which means that the upload form is currently half international copyright crash-course and half upload-interface. The plus side of this is that you can be sure as a user of Wikimedia Commons that everything there has had it’s copyright checking done for you. None of this “contact us if you would like to use the image” stuff, everything is available to use and re-use. Flickr, of course, offers a much broader range of potential copyright licenses - including non-commercial and all-rights-reserved. However, in Flickr Commons a GLAM is only allowed to use the “no known copyright restrictions” tag which means that all content in Flickr Commons is already approved by the providing institution to be used in Wikimedia Commons anyway.

3) No ads, no corporation, no commercial motivation. OK, so this one isn’t exactly new, but it’s worth reiterating. Since 2005 Flickr has been owned by one of the internet’s giant commercial enterprises - Yahoo!. Flickr Commons sits at the more altruistic end of the spectrum of their activities but the fact that Flickr is owned and operated by a US commercial entity no-doubt features as a potential risk in GLAMs meetings to assess whether to join the project (especially so for publicly-funded GLAMs outside of the US where there can be rules about domestically-sourced partners etc.). Of course where I’m going with this is that Wikimedia projects are all completely ad-free, run by a charity, charge no fees for usage, require no log-ins or personal information etc. etc. The flip-side of this is that, as a corporation, Flickr can choose to take down images if the uploader says so, the Wikimedia Foundation can’t. I’ve heard that some GLAMs have been reticent to upload to Wikimedia Commons out of the fear that they can’t delete them later if they change their mind.

4) Contextualisation. The most obvious difference between Flickr and Wikimedia Commons is that Flickr is a website for photographs to be seen in-and-of-themselves whereas on Wikimedia the images are (at least ostensibly) intended to be used in an encyclopedia. Of course there’s no obligation that an image uploaded to Wikimedia Commons would ever be used in a Wikipedia article but that is the general idea. Flickr is good for discussing photograhy as an artform in dialogical fashion (a very valid activity - don’t get me wrong) and the audience there is allowed to curate galleries quite easily. On the other hand Wikimedia Commons is good for being able to take a more curatorial approach - to embed the images in an educational context where the cultural significance of the subject/medium/author etc. can be elaborated. Both are useful things but Flickr can be a bit of an ‘echo chamber’ - especially when it’s an image of a collection item.

5) Usage checking. If you look down the bottom of the page for any image in Wikimedia Commons you will be able to see a section entitled “File Usage on Other Wikis”. This global checker is relatively new and enables you to see just how and where any individual image is being contextualised in articles across all the different language editions of Wikipedia. For example, check the usage of this image of former German Chancellor Konrad Adenauer (donated to Wikimedia by the German Federal Archives). You can see that it is used in three articles in the English edition but also two articles in Hebrew, two in Arabic, etc. etc. That’s the kind of statistical usage-proof that makes for great executive summaries to management.

5.1) Usage Checking - categories! This one is really new. Not only can you look up the stats for an individual image but now you can do it for a whole category using the “GLAMerous tool” by Magnus Manske. Try one of the “popular groups” to give it a go. This tool will aggregate the usage statistics for any category - most especially things like “category:images from xyz museum”. This lets you see in short order the combined multimedia contribution and usage of any GLAM on Wikipedia. Very nice!

Ultimately, they’re related projects with similar aims - the publication of GLAM multimedia content to a wider audience - but they go about their work in deliberately different ways. 2010 will no doubt prove to be an interesting year for multimedia in Wikimedia projects.

[update: Mia's blogpost about this topic now includes a collection of the tweet replies she received to the question "has anyone done audience research into why museums prefer Flickr to Wikimedia commons?"

Some of the responses included:

Nick: Flickr lets you choose CC non-commercial licenses, whereas Wikimedia Commons needs to permit potential commercial use?

Janet: Apart fr better & clear CC licence info, like Flickr Galleries that can be made by all! [and] What I implied but didn’t say before: Flickr provides online space for dialogue about and with images.

Richard: Flickr is so much easier to view and search than WM. Commons, and of course easier to upload.

Hopefully, I’ve adequately addressed these comments in the body of my post. iane15 had this to say in the comments:

At Hampshire County Council, the Museums Service got 99% to a Flickr Commons agreement, then Flickr said they ” need to delay adding more Commons partners until later in the year”. That was June 2009. Emails in December have gone unanswered. I don’t think we’re even going to bother any more.

Intriguing.]

[Update 2: Seb Chan from the Powerhouse Museum has just made a detailed reply to this post detailing what advantages the Powerhouse saw (and still sees) in Flickr Commons over Wikimedia Commons. Whilst my blogpost identifies what I see as Wikimedia's advantages for GLAMs, I must admit I do agree with his assessment of Flickr's relative strengths. The kicker is this:

Whilst Wikipedia and Wikimedia are, in themselves, exciting projects, their structure, design and combative social norms do not currently make them the friendly or the protected space that museums tend to be comfortable operating in.

He also reiterates the importance of the Multimedia Usability initiative which might be able to address some of Seb's points (though not all, as some are social rather than technical issues) and hopefully make Wikimedia a little bit more GLAM-friendly.]

As many of you know, until recently I was employed at the Dictionary of Sydney as the Multimedia Coordinator. I left a couple of months ago and took up a short-term contract at AustLII running the Australasian Legal Scholarship Library. However it was at the Dictionary of Sydney that I ‘cut my teeth’ in copyright and also in GLAM relation so it’s fair to say that I still have a strong connection with the project.

Simply put, the Dictionary of Sydney (DoS) is a free-access, digital history of the city - its people, stories, places, events - managed by the DoS Trust funded by the Australian Research Council. And it is a professional history project, not the Yellow Pages…

…and it just recently launched!

All DoS texts are original research by known scholars of the topic and most - and this is the bit that I’m most proud of - are licensed under the Creative-Commons Attribution Share-Alike license (cc-by-sa) and are therefore Wikipedia-compatible Free Cultural Works. All of the contributing authors were given the option of allowing their work to be re-usable and most chose to do so. This kind of optional CC licensing is AFAIK up-there as CC best practice and it was discussed in the CC-Australia blog and also in their Australasian case-studies book. You can see all of these articles by clicking “sort by license” here - hundreds of them!

Differences from, and relationship with, Wikipedia

Of course, one of the most frequently asked questions is why do we need a new encyclopedia in this era of Wikipedia. Indeed, that’s one of the reasons I was brought onboard the project - to make sure that the two projects were complementary and not competing.

1) anti-NOR
One thing that needs noting is that DoS is all Original Research: the scholarship is new; it has named authors; it has an authorial point of view. Also, unlike most professional encyclopedia, it cites its references. Because of all this DoS is a fantastic source of references for Wikipedia. DoS already links to Wikipedia in the “external links” section of some of its records about people, for example the famous photographer Harold Cazneaux or the convict Esther Abrahams.

2) Records
If you go the Wikipedia page about the Sydney Opera House you are taken straight to the article. In DoS, you are taken first to the record view which concatenates all information about the subject including a link to the article itself. In Wikipedia parlance this is somewhere between a stub, disambiguation page and an infobox and means that DoS can have records for subjects that it knows exist, but no one has yet written an article about it.

I like to think of the record view as akin to a 21st Century library card catalogue. The article contains a full text (sometimes with curated pictures alongside) but the record view contains information such as mapping, demographics, timelines, multimedia galleries and semantic relationship statuses.

3) Semantic relationship statuses
Say what? [warning! somewhat technical]
What this means is that all records are linked to each other through a series of structured relationships. In Wikipedia we have a folksonomy of categories - whatever seems to work best, that’s what Wikipedia creates. By contrast, in DoS there is a structured ontology (with relatively shallow nested depth) of types of things that any subject can “be”. If it is “sub-type: animal” then it must also be “type: natural” - see for yourself by sorting by type in any of the browse buttons on the right hand column’s toolbar.

Furthermore, all relationships between subjects are also chosen from an equally highly structured ontology. For example, the famous colonial Sydney architect Francis Greenway designed the equally famous Sydney building the Hyde Park Barracks. The relationship of Greenway’s article to the Barracks’ article is “relationship type: architect of”. This also means there is an automatic inverse relationship from the Barracks back to Greenway. There are a limited number of relationship possibilities and include things like “friend of” and “married to” and these allow you to plot the shortest distance between different subjects - a semantic Sydney-bacon number if you will. This enables the possibility for the first time to find connections between disparate aspects of the city’s history that were not previously known.

The relationships can also be given a location in time and/or place. For example, Greenway’s professional patron was the Governor of the day - Lachlan Macquarie. They have the relationship of “patron/patronised”. However, at some point the two had a big falling out and this is where the time aspect is important. This relationship was not everlasting but had specific start and end dates that can be automatically mapped on a timeline.

The relationships, and automatically generated interactive timeline of Lachlan Macquarie (whose DoS article is also cc-by-sa, by the way).

In Wikipedia there are no formal relationship statuses and therefore all links are “dumb links”. That is, the website does not know why the two articles are linked together and you have to work it out from reading the context of the linked words. Pieces of information that know their place in the database constitute the core of the “semantic web“. For the technically inclined, DoS uses “RDF triples” which is what Semantic MediaWiki and DBpedia are also working on.

[I must admit, we had good fun in the office working out what the relationship statuses would be, and especially the reverse statuses. For example, if you're allowed "friend of" can you also have "nemesis of"? And, what if the relationship isn't mutual - can you be "friended by" or "nemesis-ed of"?]

4) Essays
Most of the articles in DoS are about specific “things” - buildings, people, events, places. However, many articles are also about “subjects” such as transport, health, politics… These essays have no “record view” (described above) because they cannot be given a time, place or formal relationship status. They just are. Some are comparable to Wikipedia articles whilst others simply don’t match the manual of style for what constitutes a Wikipedia article. The list of all these essays can be found under the heading “sort by type > Thematic entries“. Some of the more esoteric essays are:
- Reading the Roads a history of road markings in Sydney, official and user-generated (cc-by-sa)
- Aboriginal Migration to Sydney since WWII which is pretty self explanatory, if complex.
- Coal Lumpers the wonderful profession of hauling coal on and off ships (cc-by-sa)

Looking over Miller’s Point, c1875-85, where Coal Lumpers would live during the week near the shipyards [used to illustrate the Coal lumpers article]

5) Anti-NPOV
The structure of the website allows for multiple, potentially conflicting, stories to be written about the same topic whereas in Wikipedia these stories must be merged into one neutral narrative. The articles do not attempt to have a Neutral Point of View. Currently there are no “double articles” of this type, but they will come in the future.

6) Scope
Obviously, being the Dictionary of Sydney (albeit the greater Sydney region) there is a geographical constraint of scope that Wikipedia does not have. This means, for example, that the article on the Chinese is only about their experience and impact on Sydney - not worldwide. Perhaps in the future Wikipedia might also include ethnographic histories at this level of granularity but currently it does not.

mmmm…. Sheep’s tongue for eight penny ha’penny and good sperm candles a bargain at five penny ha’penny per pound! [used to illustrate the Chinese in Sydney article]

Future releases of the website will be including things like:
- Mobile version, integrated with QR codes (or similar) on the official information panels around the city.
- More articles (obviously), but more importantly, contesting articles about the same subject.
- More external links from DoS out to Wikipedia articles, including links to articles in non-English editions (when applicable).

I have listed a couple of DoS’s cc-by-sa articles in the external links of some of Wikipedia’s articles: The suburb Surry Hills (WP, DoS); The Archibald Fountain (WP, DoS) and Sydney’s Trams (WP, DoS). I’ve also notified Wikiproject:Sydney and indicated my clear CoI. If these links are positively received I will progressively add some more and hopefully people will start to incorporate some of the Dictionary of Sydney’s research into Wikipedia too!

Here is the introductory article from the front page of the first edition of the first newspaper in the Australian colony - The Sydney Gazette March 5, 1803. As the indefatigable director of the National Library of Australia’s “Australian Newspapers” digitisation project (ANDP), Rose Holley, said in her presentation at GLAM-WIKI - this is a manifesto that the Wikimedia movement would feel strong ties to. I can imagine that if the publishers of that newspaper had been alive today, they might well have been Wikimedians - and vice versa.

http://newspapers.nla.gov.au/ndp/del/article/625438

[Innumerable as the Obstacles were which threatened to oppose our Undertaking, yet we are happy to affirm that they were not insurmountable, however difficult the task before us.
The utility of a PAPER in the COLONY, as it must open a source of solid information, will we hope, be universally felt and acknowledged, We have courted the assistance of the INGENIOUS and INTELLIGENT :--- We open no channel to Political Discussion, or Personal Animadversion :--- Information is our only purpose; that accomplished, we shall consider that we have done our duty, in an exertion to merit the Approbation of the PUBLIC, and to secure a liberal Patronage to the SYDNEY GAZETTE. ]

Replace “political discussion with [[WP:SOAP]] and “personal animadaversion” with [[WP:NPA]] and you’ve pretty much got the essence of what the Wikimedia movement is all about. We’re even starting to “secure liberal patronage” through donations like the one announced today from the Hewlett Foundation!

I’d like to also point out how absolutely amazingly cool the “Australian Newspapers” project is. You might notice that the link given to find that image above is specific not just to the newspaper, not just to the daily edition, not just to the page - but specific to the article! They have made a persistent and stable URL to every single article in every single newspaper edition they have. Furthermore, all the newspapers they have scanned are in the public domain and have a “save as PDF” “save as picture” and “print” function. That’s how I made the above image. Too easy.

Furthermore - notice that they encourage the public to correct the text. That’s right - a national cultural institution that’s not afraid of asking the public for their help - and they have the most stunning statistics about how much (and how well) the public is helping. They even have their own hall of fame for the most prolific text-correctors.

Finally, if you go to the “about” page for any of their scanned newspapers you’ll see that they’re live-linking the first few sentences from the respective Wikipedia entry. See the page for the Melbourne Argus here for example. This is a fantastic use of what Wikipedia does best - simple descriptions of specific things. It’s not the National Library’s job to write descriptive passages for every single newspaper in Australia’s history - it’s ours.

Here are all the specs, usage stats, workflows, ORC and Metadata info, system architecture, CMS, correction system structure… http://www.nla.gov.au/ndp/project_detail and here’s their title availability info with a link to the tantilising “titles coming soon” list.

In the lead up to the (now fully subscribed!) GLAM-WIKI event there is an increasing amount of chatter about how can Wikipedia play a part in helping the cultural sector to get their knowledge out to the world. But I’ve just come across another area that we’ve barely scratched the surface of:

Local history.

(Bachman farmstead workers load produce onto a Dan Patch line boxcar for delivery to market. Richfield Minnesota - date not published )

Every city, town and community has one. A small group of people who try to put together their photos and memories about the place where they live. Often this gets produced into a locally made coffee table book, often this work lies dormant.

But this is where Wikipedia and Wikimedia Commons can come in great use. Take, for example, the article [[History of Richfield, Minnesota]]. This is largely written by [[User:Richfieldhistoricalsociety]].

Here’s the story (at the “Museum 3.0″ Ning) that put me on to it:

“I can tell you from a small museum perspective Wikipedia is invaluable. I am also on the board of the a tiny historical society in near Minneapolis, Minnesota. I helped them set up a wiki for them to collect and disseminate the history of our community. However, not only did no one contribute - no one bothered even to go to it. Instead, we found that putting our community’s (fascinating) history on Wikipedia reached more eyeballs than we ever could if we left the information ghettoized on our own now defunct wiki.”

- Joe Hoover from the Minnesota Historical Society (on twitter @nyargle and blogging at nyargle.com which is subtitled “figuring out how to market the museum on the web one pixel at the time”)

I find the word “ghettoised” particularly interesting. It raises the very true point that whilst so much work goes on around the world in getting local history written and published, it is largely kept apart from the rest of the world’s collection of knowledge. Isolated. Often this is due to the expense of mainstream publication and the narrowness of the subject matter. But this is exactly where Wikipedia can help! We can host those town’s histories (so long as there are verifiable sources, e.g. the aforementioned coffee-table book) and bring them out of the isolation and stagnation described by Joe.

So what can we do to encourage this more?

1) Let me answer my question with another question:
If you were asked to go along to your local history society meeting and give them a practical training session on editing Wikipedia (even assuming they had access to enough computers) would you either say “yes, I’d love to spend the next 3 months teaching you how to read WikiMarkup” or would you say “Editing Wikipedia is quite complicated and I think I don’t have the time to help your local studies group get up to speed.” Just thinking of how tricky it is to edit tables, add references, explain nested templates, upload images, decrypt infoboxes… gives me the heeby-jeebies.

Editing in MediaWiki is just too damn hard for the majority of the population. It makes them feel stupid and frustrated. The usability team will be able to make this learning curve less steep but I don’t think they have the resources or the time to do as much as is clearly needed within the scope of their current grant. The less steep we can make the learning curve to independent editing of Wikipedia the more likely that different interest groups - especially those that are not particularly technologically inclined - will be able to join in.

2) Localisation support. As Gerard Meijssen often reminds us, and is the reason for his standing for the Foundation board, we have a very haphazard approach to supporting languages other than English. The “in your own language” part of our vision statement is not given much financial support or attention (relative to the “for free” part for example) that it deserves. I’m not sure how this should be fixed but it certainly needs to be addressed. The local histories in the English speaking world are important enough, but imagine how interesting and diverse the local histories from non-English speaking areas of the world are!

(Housing development along Washburn Avenue, Richfield Minnesota, 1950s)

I just returned from the 2009 Australian Historical Association conference, held this year at the University of the Sunshine Coast. Lovely name, lovely place! Can’t beat a uni campus with wild kangaroos running through it…

I’ve given copyright presentations to lawyers, and ‘freedom’ presentations to linux folk, but this was the first time I’ve given a history presentation to historians. And, since I’ve just graduated in history, these were the people I had to impress.

So… I chose a chapter from my thesis called “Wikipedia: the endless palimpsest” which argues that Wikipedia can be legitimately used by historians in several ways as a primary source.

Embedded is the video of the presentation and after that is the abstract. The presentation went very well in the end. I got a packed-out room and several people with some fantastic expertise came up to me afterwards and asked how they could get involved.

Wikipedia: The Endless Palimpsest from Liam Wyatt on Vimeo.

(Sorry about the poor quality, it was only filmed on a digital camera and I had no microphone).

Abstract:

The palimpsest is that most unusual of sources as it shows not only the final state of the page but allows us to see what is now valued but was once discarded. The wiki methodology of writing, with its inherent ability to return, compare and restore to previous versions of any page can therefore be seen as an infinite palimpsest—digital vellum being scraped back, written over and restored ad infinitum. Just as Pompeiian graffiti is of interest to academics two millennia later, so might some otherwise unprepossessing text in Wikipedia’s archive be of interest to the future’s linguist, historian or sociologist.
The ability to see snapshots of articles at any given point through history, their corresponding discussion pages, associated paratexts and statistics demonstrating article popularity gives Wikipedia great potential for historians. This paper uses these elements of Wikipedia to highlight practical means by which historians might engage with it as a primary source of history and still maintain professional standards. Therefore, under discussion is Wikipedia’s own historical record and how it could be used to great effect by historians.