The Saint's Birthday Message

It is my birthday today! Not my real birthday, but my official Saint's Day: it was a year ago today that I adopted the moniker @SaintEvelin, and began to embrace and become part of the online library community that had daunted me until that point. So it is that I have been attempting to succent without sneers for a whole twelve months, and in this 43rd blog post I shall gaze at my belly-button to see if the fluff therein has corresponded to the aims I set out in my first post, or whether, like Charles Foster Kane, I have diverged from my declared principles.

"This blog...will document my attempts to find my place in, and embrace firmly, the library community..."

Of the 42 posts prior to this one, 33 (79%) have been about libraries and/or my personal or professional development. This figure includes 16 posts devoted to the CPD23 project. Beyond this core, five posts have been about language or literature, and four posts have been concerned with that other great love of my life: fashion. I think this is a reasonable division. If one removes the crutch of CPD23, 65% (17) of my posts have been 'on-topic'.

"The title of this blog: "Succentorship without Sneers", describes (with pithiness and alliteration) my desire to balance a passion for libraries with an inherent suspicion of almost everything. This is not to say that an empirically minded doubt of all is a bad thing; it is demonstrably not. But I am a pessimistic soul at times and all too often place barriers that are unwarranted. This bloody-mindedness must stop. Thankfully I am an optimist at times too. This isn't a question of faith or doubt, but of genuine openness: of listening and considering things without taint of preconception. It's also about showing a bit of enthusiasm and not being such a miserable git: beyond any question of science or discrimination, there is a simple matter of attitude: I have to go out and get stuck in, rather than stay sitting here on my beautiful arse, waiting for the world to come knocking for my services."

Here's the nub of the blog. It's not just meant to be me sat here typing my thoughts about libraries. It's supposed to be me getting out and engaging with the profession; expanding my horizons and embracing new developments. I don't suppose it's helped that I've still not got myself a professional post: it doesn't encourage an awful lot of optimism and it leaves me with only so many new and exciting frontiers to explore. No surprise then that this blog has descended into the odd didactic rant now and again (I am still, it seems, "a cynical bitch"). But I've had three Library Camps to help inspire me along the way, and to have had a blog post referenced in print has been a particular honour: a real validation that I must be doing something right here. Above all, my new-found library friends both in the flesh and on Twitter have been invaluable in keeping me perky and engaged during the discouraging tedium of job-hunting. It really means a lot.

"I shall attempt to be honest both to myself and to my readership, whilst maintaining a reasonable modicum of diplomacy."

I do think I've broadly succeeded in this regard, though I think I need to do more reflection. I'm a timid creature, and I'm always a bit scared about how my attitudes and opinions are received by my peers (all too often I've snuck my posts out when nobody's looking). I need to be braver and I also need to be more honest and open. That said, it's hard to be fully open without breaking into distracting exhibitionism; there's always an element of fear in expressing oneself, and I suspect that there's a happy medium to be achieved somewhere between retention and extroversion. Locating this medium is something I find a considerable struggle in reality let alone online. I'm always concerned, for instance, that people find it irritating, weird or disturbing when I tweet about pretty clothes, just as I'm concerned that people find it irritating, weird or disturbing when I wear them. So it is that I try to limit my tweets on such topics and avoid allowing my every blog post to descend into some hand-wringing whinge about the social tribulations of a genderqueer librarian. As much as this aspect of my identity might impact upon my personality and my interaction with the world (and it is, for me, a genuinely big issue with which I'm constantly wrestling as I take these first steps in my career as a librarian), I endeavour to treat it as something unworthy of special attention or remark: to normalize it rather than to have this blog become dominated by that theme. I don't really want to find myself pigeon-holed in such a way. Perhaps this is the wrong approach to take. Perhaps it's unhelpful, or maybe even fundamentally dishonest. That's a big worry I have: that I mislead through omission and give an offensively false impression to my audience. I don't know. It seems a difficult balance to strike. I welcome any thoughts you might have on the matter (and on anything else I've had to say today).

Aims, then, for my second year? Be less timid, but not too ranty. Keep sharing my experiences and what I've learnt. Hopefully I'll be in a new job soon, and then I shall have a world of new learning to exploit, but there's still much I can write about even now (I'm currently working on some hopefully helpful tips for Excel, in response to a suggestion on Twitter). I feel I've come a fair way in the last twelve months: no longer am I a terrified and daunted newbie. I'm increasingly engaged with Twitter, and thanks to the Library Camps I'm starting to get to know people in real life too. It amuses me a little to think that just over a year ago I approached my first library conference with real dread, while now (thanks to the near-damascean conversion I had there, and the warmth and encouragement of the library community I've entered) I'm filled with excited anticipation for the CDG event in Birmingham next month. If you're going, I look forward to seeing you there, and whether you're going or not, I'd just like to take this opportunity to say:

...thanks for reading. x

The Great Sheffield Book Hunt

In my last post I talked a bit about catalogues and how I should write a post in which I test-drive a few. So here we are. I've given five Sheffield-based library catalogues a spin: two from the two Universities in the city, and one from the public library service. Let's take a look at tonight's contestants:

Sheffield Hallam University's Innovative WebPAC Pro "Classic" catalogue:
The out-of-the-box OPAC for the Millennium LMS. This is a simple little piece of kit with a drop-down choice between searches for Keyword, Author, Title, and a few other bits and bobs (plus an advanced search option). But for the challenges we're about to put its way, we shall only be using the default Keyword option.

Sheffield Hallam University's ProQuest/SerialsSolutions Summon "Library Search" catalogue:
After dabbling with Innovative's "Encore" for a year, SHU have recently (last Autumn) opted instead for the Summon federated search catalogue system. While there's an advanced search option, and opportunities to limit searches to particular criteria, the default approach is a Keyword search of everything except newspaper articles and book reviews (which were taken out of the default search just a week before I conducted this test), and it is upon this default that we hurl our challenges.

The University of Sheffield's Talis Prism "Star" catalogue:
Upgraded from a previous Talis system in 2003, Star definitely looks its age. There's a choice of Keyword, Author and Title search, and again we shall be taking the Keyword option for purposes of parity.


The University of Sheffield's ExLibris Primo "StarPlus" catalogue:
They thought it would never happen, but at long last Star has been updated. The new catalogue (in a perhaps perpetual "Beta" stage) came on-line in November 2011 (not long after SHU's Summon). The default search is by Keyword and limited to university collections (including journals and etext subscriptions), with a secondary option for literature cross-searching across multiple databases. There is also an advanced search, but we shall be using the default set-up.

Sheffield Public Libraries' SirsiDynix iBistro library catalogue:
The aesthetics of Star alongside the drop-down functionality of the classic SHU. As with the other contenders, we shall be using the Keyword ("words or phrase") option for our search experiments.


Our contestants met, it's time to start...

Round One: Vital Statistics

"Clicks to search" is the number of clicks it takes to reach the catalogue search box from the institution's front page. "Results per page" is the number of search results per page. "Average loading time" is based on a search for pride and prejudice and is the mean number of seconds calculated from five successive attempts on my home connection of a Sunday afternoon. The results of that search are converted to MHTML to give the page size (in megabytes).

Both SHU catalogues take quite some finding from the University's homepage, although access is more direct from the student homespace and VLE. The slowest loading time was achieved by the public library, despite having the second smallest page size, but such is the nature of the server. Heaviest on the bandwidth is Summon: its results exceed the 1MB mark, making for a loading time twice that of its classic companion (the fastest of the five on the day) despite the fact that the classic search packs in twice the number of results per page. Unquestionably, then, SHU Classic is the most efficient in terms of time and bandwidth.


 Round Two: Error Correction

Just a quickie, to test the catalogue's error-correction capacity. SHU Classic and Star don't do correction. What you type is what you get, and what I've typed is pride and predujice. The public library offers a list of "closest matches" which includes what we're after (the novel by Jane Austen), so that's a pass. StarPlus correctly deduces that we're after pride and prejudice and offers that as an alternative search, so that's a pass too. SHU Summon, however, thinks we want pride and predjudice which gets us 28 (irrelevant) results. Only then does it offer pride and prejudice.

So having finally got some results for pride and prejudice, how do they look?

Round Three: The Austen Test

Here's three formulations of the title: pride and prejudice, pride & prejudice, and "pride and prejudice". The first is a basic test, the second will challenge the catalogue's ability to cope with ampersands, and the third will demonstrate the effectiveness of quotes.

Star does not do ampersands (not in a Keyword search at any rate). It cries if you try.

In the above table, the second column for each catalogue is the number of matches retrieved, and the first column indicates how far down the list we have to go before we find our first example of the Austen novel. With the exception of the public library, our book appears on the first page of results, and in the case of StarPlus, it is right at the top. Curiously, although it limits the number of results, adding quotes to the search also drops the book lower down the page for both SHU catalogues.

The lower number of returns for StarPlus compared to Star is down the way that StarPlus groups like titles within a combined record. The higher returns on Summon compared to SHU Classic are on account of the breadth of external material being included within the search.

Round Four: The Proust Test

In this round we test the way the catalogue (and perhaps to a larger extent the cataloguer) handles the questions of translation and the episodic novel. In all seven of the above searches we're trying to find for ourselves a copy of Proust's "Swann's Way" (in either French or English; we're not fussy). The diacritics in the French formulations garner no results from the public library and the classic SHU catalogues, which is unfortunate. In the case of the SHU catalogue, at least, this is due to the diacritics being absent within the catalogue records. The University of Sheffield has superior catalogue records regarding its use of alternate titles, but the search also seems to strip diacriticals from a character, thereby giving the same hits for Du côté de chez Swann as it does for Du cote de chez Swann. On the first three searches, Summon fails to find our book within its first 100 records (an arbitrary cut-off on my part) though does manage to find later volumes. The Star catalogues on the other hand are immensely successful in giving us what we're after, with StarPlus being spot-on every time.

Round Five: The Godard Test

Here's a similar test, but one which is broadly passed. This time we're looking for a film by Jean-Luc Godard. The only failure is the missing diacritics in the public library, but the grave has made it to the SHU catalogue record this time round. SHU has a problematic lack of overlap, though, unlike Star's results.

Round Six: The Taylor Test

Now we shall see how the catalogues cope with a popular initialism. Ideally, the results for ajp taylor should be a subset of those for alan john percivale taylor, and this appears as if it may be the case for all the catalogues except Summon, where the noise from various articular references to AJP creeps in. Understandably, the a j p taylor returns are comparatively high, but in all cases bar Summon we still get a top rank result (we don't get a work by AJP Taylor till page 2 on Summon).

Round Seven: The Orwell Test

This round is another test of the ability to locate a work given the author's name. We're after any individually credited work by George Orwell (letters excluded). Unsurprisingly for a Keyword search, the search for george orwell gives the same results as the search for orwell, george, and likewise with g orwell and orwell, g. The problem with searching for the author by Keyword is that there are an awful lot of books about the author, with the author's name in the title, and these tend to rise to the top of the rankings. Summon is a particular victim of this, and it takes a set of inverted commas to bring any works by Orwell into the top 100 returns. Conversely, adding the quotes to StarPlus drops our first hit beyond the first page. In fact it is the old version of Star that is the most successful in this round (failures on the initialised formulation notwithstanding), while only SHU Classic gives us a front page hit for every case.

Round Eight: The Plural Test

I want to find a book about search engines. Or should I find a book about search engine stuff? A wildcard would go well here, but is unnecessary in Star which automatically knocks off any terminal "s" and adds those findings to the pool. So it is that a search for foxes in Star brings up a page of works by John Foxe. A similar phenomenon appears to be occurring in SHU Summon, although it seems the results are more relevantly ranked without the plural than with it.

Round Nine: The Amey Test

Our contestants have made it to the final round, and the toughest of them all. For this is the specialist round. Their task is to find "The Collapse of the Dale Dyke Dam, 1864" by Geoffrey Amey. It's a book about the Great Sheffield Flood of 1864, and it's a book that all three collections contain. To find it we shall use the simple but likely search terms sheffield flood and sheffield floods. As can be seen from the above table, it was not a task well-met. Again, the problem here is one of cataloguing more than it is of the catalogue itself. Even the one catalogue to give a direct hit (thanks to the MARC 650 entry "Floods|xEngland|xSheffield.") does so in a rather hit and miss way (three of the ten hits for the sheffield flood search are relevant texts, with only one of them appearing in the results for sheffield floods . Star uses its plural trick to give us all the books it can in both cases, but Amey is not listed among them (despite being in the collection). In Summon it's the last book we deign to look at, right at the bottom of page 4, and Sheffield Public Libraries aren't much more forthcoming. But at least it is there in both cases.

Final Results

Summon and StarPlus consistently returned results irrespective of the rigour of our tests, though Summon gave an average return of 58,203 results which is three orders of magnitude out from the rest of the catalogues. Such high returns were not stacked in our favour, however, and the average relevance ranking of 23 is the lowest of the five, though it is at least on the first page of results. Most successful in retrieving our desired items (and also in ranking them prominently) is StarPlus, though Star and SHU Classic don't do badly in these regards either.

The above tests are all simple searches made upon the default search settings of each catalogue. The noisy irrelevance we encounter in Summon could be cut back were we to refine our search, either through the advanced options or via the limiters in the side menu. Unfortunately, the bulkiness of Summon makes it the slowest of the four Higher Education systems here, and we know from Google how important speed is in our searches. Having to make secondary moves, such as limiting our searches to books, makes the search process all the slower. Those behind Summon would point out that I am using a screwdriver to hammer a nail here; that Summon is geared towards the literature searching needs of the modern student, and is more than a book-locating tool. But I do feel that book location is an essential feature of a library catalogue, and that Summon isn't really all that good at it. Again, I throw in the caveat that we've been using Keyword searches where more specific advanced field searching might serve us better, but the same caveat applies across the board. When StarPlus is capable of being near-consistently on the money for us, while also offering a range of federated search options beyond the main holdings, the bulkiness and noisiness of Summon appears unhelpful.

But it isn't all about the catalogue system. As a number of the examples demonstrate, the catalogue record itself is of immense importance. Charles A. Cutter wrote (in 1876) that the purpose of the catalogue is "To enable a person to find a book of which either (a) the author / (b) the title / (c) the subject is known... To show what the library has... [and] to assist in the choice of a book". The most difficult index to achieve is that of the subject, which is why we have things like AACR2. Attempts to Google-ize the catalogue experience have all-too-often fallen at this hurdle: it's hard to build a foolproof relevance ranking from MARC fields, as some of the above experiments show, and things can get particularly messy if you begin to introduce some full texts into the mix. This is what seems to be happening in Summon: full-text articles are getting more matches for the search-term and are rising to the top of the results at the expense of the terser records. In the long term, when everything in our collection is fully searchable, such Googly games might actually work, but we're not there yet. That's not to say that one couldn't come up with a sensible (and probably quite complicated) algorithm to deal with both types of material, but Summon has not done this, and StarPlus has gone with a less integrated model in an effort to avoid such pitfalls. On the strength of the above tests, it's an approach that seems to have worked: as a catalogue of the university's holdings it is undoubtedly an effective tool. By not attempting to be a jack of all trades, StarPlus succeeds in mastering the core role of the catalogue. Whether it operates as successfully when employed as a wider literature-searching tool remains to be seen. That's one for another day's testing.

Where is Horsforth? Library Camp Leeds

I don't like Leeds. It's a prejudice formed in part from the Leeds-centric bias of regional news magazines, in part from the city's dingy train station (the first at which I encountered the evil that is ticketed entry and exit), and in part from an ill-prepared visit there that failed to locate any decent pubs. But mainly it stems from one cold, wintry night when I had to walk the streets of Leeds till dawn's first trains, and in the process (thanks to recent flooding) managed to tread on a fish.

But Library Camp Leeds was not a city-centre affair. Library Camp Leeds was in lovely, leafy Horsforth: a land uncontaminated by my fishy prejudice. In Horsforth the sun shines brightly on visiting librarians.

It was with some nervousness that, in the weeks ahead of the event, I noticed a clash in my diary. For Saturday 26th May also held a sacred and annually observed holiday: The Eurovision Song Contest. Fortunately, public transport would allow me to get home in time for the hallowed event and still have opportunity for a post-Camp drink.

In tribute to whatever higher powers permitted this glorious conjunction, I decided to bake for Horsforth my first Library Camp offering, settling on a Honey Flapjack adapted from this BBC recipe. I'm glad to say that it went down well (although perhaps not as well as @daceudre's cheese sticks). 

With the sun now promenading in full 'thermal death ray' mode after several months in hiding, I had to make a last minute change to my intended wardrobe in a desperate attempt not to melt. This sudden warmth in the world brought out on show a splendid array of sartorial summeriness (high on my want list was @bumsonseats' rosy number), and we all looked rather wonderful in this rare sunshine spotlight.

But we're not here to talk about cooking and clothes... well not just about cooking and clothes... there's these funny things called libraries to discuss. As ever, my descriptions below use the sessions as starting points for a few thoughts of my own, but, where available, I've linked to other accounts of the day.

Session 1: eBooks

@pennyb led this session, and you can read her account of it at LISNPN. I started writing a post on this subject a few months ago but it's such a knotty, frustrating and fast-moving area that I never finished the article and it's now looking somewhat out of date as it languishes in my drafts pile. Nonetheless, the ebook question is probably one of the most important issues there is in libraries at the moment, and the ©-shaped knots involved go right to the core of what libraries are. At some point in the future, when all monitors use coloured e-ink and when wi-fi is available to us all wherever we may be, the ebook question will be somewhat easier and more straight-forward to negotiate, but that future is a fair way off and in the meantime we have to deal with a mess. We currently inhabit a world where Kindle has become the Hoover of e-readers, and the end of the Amazon-Waterstones war will make this state of affairs all the more true. I think we can thank lazy journalism for this Kindle hegemony: certainly, in a Which?-style comparison, the Kindle's limited support of file-types makes it look as enticing as a hi-fi that consists only of a radio, when I've got a stack of CDs, tapes and records I want to play. This lack of support extends to library ebooks, and one can imagine the frustration of the newly Kindled library member turning up to borrow an ebook and being told they've bought a dud. But the frustrations are not limited to Kindle-owners. Even a library member savvy enough to have bought an e-reader that can read the file-types their library offers will be faced with an array of irritating limitations imposed by the terrified publishers-that-be. The situation is farcical and far from user-friendly, although that is part of the point: publishers very much want the path of least resistance to be the one which makes them the most money, and at the moment it's the publishers laying the terms. It's hard to see this mess being overcome without some collective endeavour on the part of libraries, and ideally with the help of forward-thinking government legislation (a fantasy in the current political wilderness).

Underlying all of this is the Big Question: the internet-created challenge to the economic model that has sustained the creative arts since the advent of mass-media. How do we fund the arts?

When libraries started out, professional authors did not really exist. Even today they constitute a small minority of writers, with PLR offering token recompense to the rest. With a fixed print-run, a publisher's concern is that all the books are sold, and in some cases a library in every town may benefit those sales. The lending of books was no great threat, plus it educated the masses and increased the market for future works, creating a world where books can now fill a whole aisle of Tescos. eBooks are more troublesome: there's no PLR, and the ephemeral nature of the medium in the eyes of the market makes for a retail mark-up with minimal margins. When anybody can stick a pdf on the internet, the role of the publisher is reduced to finance, quality control and distribution (something which the library itself could conceivably facilitate). The loss of the physical book (and its appeal as an owned artefact) means that a bought ebook is practically no different to a loaned ebook: the two are potentially as accessible, and both are in identical condition. If libraries are able to loan ebooks there becomes no reason to ever buy an ebook. It is for this reason that publishers are seeking to make the process of ebook lending such a faff: to make it far easier to buy than to borrow.

Perhaps the solution is to address this equivalence problem in a different way. Not all publishers are so troublesome to libraries, and in the realm of magazines there are some genuinely good deals to be had (Zinio, for example). Why should magazine publishers be so much more friendly to us? Because their funding model is built largely upon advertising, and the more people who see those adverts, the higher the advertising fees can be. Perhaps, then, a future model for ebook loans would see library ebooks subsidized through advertising, much like the free version of Spotify.

Session 2: Can we be funny?

@llordllama asked this question, primarily with respect to how we present our teaching sessions. The answer was a resounding "Yes" (such were the acoustics). There then followed a few anecdotes regarding how best to thaw an icy audience. I must confess that I have in my time been one of the stonier students: when faced with a Tiggerish presentation I tend to assume the Eeyore position. Such are the complexities of humour (as demonstrated by the success of Michael McIntyre): one person's joke is another's cringe. In that respect it's important to not lose sight of the aim of a teaching session: to impart knowledge. The way in which we go about imparting that knowledge relies very much upon ourselves. Michael McIntyre might make me despair for humanity, but I am none-the-less aware that he has something called a "Man Drawer" where he keeps the sort of things late night 'phone-in competition organizers store in their handbags. This knowledge has reached me irrespective of my enjoyment of the delivery, and while I may not have been amused, it was evident that several others were, himself included. This latter point may be the crux of the matter: a significant part of why I use humour in my teaching is as a crutch for myself: as a bolster against my shyness and insecurities; any thawing or engaging effect is a bit of a bonus really.

Session 3: The anti-social catalogue

The crew of the Liberator take delivery of their shiny new OPAC.

It being such a lovely day, we decided to head out into Horsforth Hall Park in the afternoon, to bask in the dappled shade of trees and to fend off the interests of passing hounds. There, in a gentle rain of blossom, @preater led a discussion about 'next-generation library catalogues' and whether their efforts to integrate aspects of social media are successful. You can read his account of the session at his "Ginformation Systems" blog.

Catalogue search strategies are a big interest of mine (they were the subject of my dissertation), and I analysed the retrieval characteristics of a range of catalogues as part of an information retrieval project at library school (I should really update the results and write a post on the subject). This latter exercise was sufficient to instil within me a great disaffection for the modern, Google-influenced trend of having a single search box at the front of a catalogue. Such an approach would be fine, were the results to be derived from the full texts of all the items available to the library. Alas we are not quite there yet, and a keyword search of a catalogue record can throw up some pretty wild results. A keyword search of "George Orwell" for instance, is unlikely to find a work by the chap until about a page into the results, because the metadata in a biography or critique of Orwell will tend to throw up more hits for his name.

This is a superficial problem which can be overcome through search refinement forms or advanced search options, but such moves take time. The thing that made Google such a success was not the simplicity of its design so much as the speed at which it returned its (relevant) results. A significant further problem with many of the next-gen catalogues I've encountered is that they are bandwidth and memory hungry, splattered as they are with flashy scripts and third-party content, and that all this has the unfortunate tendency to make using such catalogues a slow and laborious process. This process is made all the more sluggish when the catalogue is being used on a dedicated terminal, which may well be an old computer and one which gets little exercise.

The new generation of catalogues do have some useful tricks up their sleeves, though, perhaps the most beneficial being their error correction and "Did you mean...?" suggestions. Amazon-inspired features such as reviews, related reading details and user-generated tagging are of more questionable benefit, and seem not to be particularly well used. Perhaps our members have better things to do than to write content for our catalogue. Third-party reviews drawn from a broader pool may stimulate interest, however, and there's a wealth of material hiding in our LMS that might be utilised to illustrate the value or popularity of a title. The catalogue in use where I work displays the last return-date of an item, and this in itself may serve as a basic indicator of currency. One might expand this into a full date sheet for the book; it might even be formatted using a bespoke date-stamp font for added nostalgia! Our dataset is our oyster in such things, at a time when information is considered beautiful. The only drawback is that we might further fuel demand for our most popular hotcakes, rodding our own backs in the process.

Session 4: Space

@bumsonseats kicked us off on this final session, a continuation of a theme encountered at the Manchester Library Camp in January. Her summary can be found on her "Growing a library" blog. As I considered a couple of posts ago, we have to be careful with what we do to our library spaces, because we haven't got so many and we've got to use them to their maximum potential without damaging them for our core membership. It's not an easy task to negotiate, especially in the realm of public libraries, where resources are tight and spaces old and scarce. One might weed out the old and tatty stock in an effort to look clean and welcoming only to find that there's no longer much stock at all and the place looks sterile and empty. Any revision of space needs to be properly thought through, well researched and well designed, but that sort of thing takes money and that's the one thing we haven't got right now. Consequently we have to muddle through and compromise as best we can. The list of points in @bumsonseats' blog post gives an idea as to the sort of stuff we should be considering in such circumstances.

Sessions over, we went off to the pub, where I had just enough time to squeeze in a couple of glasses of Pimms before I had to get my train. I only wish I could've stayed longer and spent more time with everybody there, but alas, Charpentier was calling...

Thirty years ago, the contest was surprisingly close.