Debate: “Although digitization is useful for accessibility, detailed online item-level cataloguing is even more so”

I remember seeing that tweet from the “‘Democratising or Privileging: the Future of Access to Archives” conference at University of Dundee last week and thinking that many in the U.S. might find it a controversial statement. I was reminded of it  this morning when Jane Stevenson retweeted it and linked to Amanda Hill’s blog post summarizing the conference sessions. To put the claim in context, here is Amanda’s summary:

Chris Paton is a professional genealogist and his pleas to archivists included a request for free wi-fi in archives, permission to take digital photos, longer opening hours and simpler user registration and photocopying policies. He also thought it was important for archives to make use of social media tools like Facebook and Twitter. Both Chris and Alan emphasized that although digitization is useful for accessibility, detailed online item-level cataloguing is even more so, especially in a time of financial constraints for researchers (and everyone else!), although they both recognized that this is much harder to get funding for than ‘sexy’ digital imaging projects.

As you can see this claim is being made from the perspective of an archives user. So I put the question to you, archivists, historians, researchers and users of archives: do you agree or not? Which is more important? I expect many of you will say that both are equally important, because of course, most people want it all. But most archives can’t afford to invest in everything so, for the purposes of this debate, which is more important, digitizing materials or providing item-level cataloging?

(And if there’s some nuance to this that we’re missing on this side of the Atlantic, please enlighten us.)


Be Sociable, Share!

31 thoughts on “Debate: “Although digitization is useful for accessibility, detailed online item-level cataloguing is even more so””

  1. I think it’s important to note that he is a genealogist. In this particular realm of research, item-level cataloging (or at least a master list of names) *could* prove useful. If he was researching a specific event or subject, however, the item-level cataloging might not prove so useful, especially if that item-level description didn’t happen to suss out the particular bit of information that he was interested in.

  2. I am really torn about this because I currently work on a project which does detailed item-level cataloging in part because it originated with a group of researchers instead of archivists. Being researchers, they approached it from an user-oriented point of view of needing to find stuff and thus needing a really detailed level of cataloging to facilitated that. That wealth of legacy metadata has let us do some amazing things with our records, but it also takes an exorbitant amount of time to process items and we’re held back by lack of paid staff to do it all. If we ever plan to do real mass digitization, I’d have to cut back that level of detail to “good enough” just to get items online.

    I think ultimately it comes down to making decisions based on what resources are available and what a project is trying to achieve. For most archives doing digitization, item-level metadata doesn’t make sense because it’s time consuming when they’re trying to prioritize access and speed. But for specific projects (like digitizing an underutilized photo collection) it may be worth the time and effort.

  3. We do both at my workplace but still do item-level description if the collection warrants it. It’s my call being a lone arranger if I think it’s worth my time. Over and over again we hear from researchers that item-level is what they want (when they peruse the finding online in their jammies) and how frustrated they are with minimally described finding aids. I’ve found myself annoyed at minimally described collections too, saying to myself, “well, that doesn’t help at all” when I look at the sparse listing. Sometimes it’s best to do the work upfront, especially if you know the collection will be asked for a lot with reference questions.

  4. Comments so far from Twitter (since I’m assuming many of these people won’t post comments here):

    ‏@pashulman Cataloging also clarifies subject connections that stimulate research; helps to better prepare. Otherwise, just a black box

    @pashulman If I had a choice, item level cataloging is more useful. If it’s clear there’s a useful letter or two, I can ask for copies.

    @Flitcraft digitisation is next to useless without indexing and cataloguing. Both go together.

    @gilliandoctor Cataloging because it is hardly sensible to digitize when you don’t know what you actually have

    ‏@torstenreimer can we not have both? 😉 As a historian mixed feelings about this but surely good cataloguing helps…

    @galpix Yes, time to step back, get back to the basics of processing to provide worthwhile access, and get those online.

  5. I think good online cataloging is more useful than digitization. However, it is time consuming & many archives do not have the resources or staff to handle item-level description. I would love to give the users what they want, but if the resources aren’t there, it is near impossible.

  6. Interesting debate, but I don’t see how an either/or answer is possible. As Christine notes above, decisions are best made at the collection level: some demand item-level description, some don’t. The right balance of devoting resources to cataloging or digitization will vary with the repository; archivists must know their users’ needs in order to decide wisely.

  7. “@Flitcraft digitisation is next to useless without indexing and cataloguing. Both go together.
    @gilliandoctor Cataloging because it is hardly sensible to digitize when you don’t know what you actually have”

    I think these are both interesting points. I know there are lots of complaints about how digitization only occurs for ‘popular’ items, and ignores the bulk of collections, which are likely to not have a good finding aid either. Lack of knowledge about a collection only perpetuates this cycle. But many of us have to get grants – and it’s easier to craft an argument for working on fun, popular, or otherwise sexy collections. Item-level finding aid doesn’t scream sexy, though I do feel sure it would be way more useful for the majority of users anywhere that I have worked.

  8. A few more comments from Twitter:

    @allen_heather The cataloging vs digitization is a constant battle. Higher-ups understand pretty pictures on the internet.

    @bethknazook Strategic descriptive choices get you there, the image answers the rest.

  9. Lindsay–interesting that you echo the point Heather made on Twitter–that digitization is more popular with funders than cataloging.

    And to Rob’s point (and others) about this being a judgment call that has to be made based on understanding of users, I often wonder how well we really do understand our users. Do we understand the users best who speak the loudest? Do we value the needs of some types of researchers, who value digitization, over the needs of users like genealogists, who want more detailed cataloging? Have studies been done on how useful digitized collections are to to whom they are most valuable?

    Again, I think this is an interesting question since overall the trend in the U.S. seems to be toward less detailed descriptions. Although usually this discussed has been framed as an opposition between having more detailed descriptions of fewer collections vs. having less-detailed descriptions of more collections (aka MPLP). I have not seen it framed as a choice between more description and digitization.

  10. I think it’s good that I’m hearing a lot more discussions about what the users of our archives need. I think genealogists like detailed cataloging because they are always looking for certain people, but I also know a lot of genealogists who wish everything was digitized because the research is their hobby and they can’t afford to run all over the country looking for archival collections. Ancestry has spoiled them 🙂

    For the academic researchers, which many of us also serve, I would assume that many of them would enjoy digitization (they too lack funds), but since they are often looking for ‘new’ or obscure items, detailed finding aids are typically more helpful than digitization (at least digitization as we do it now – i.e. often only the most popular items). That is what I found myself as well, when I was doing historical research. File-level cataloging and lots of subject and keyword terms were likely good enough though – not necessarily item-level for everything. Depends on the collection, though…I guess that’s the trouble. Everything is so circumstantial depending on your collection and users.

  11. It’s hard for me to understand why this is even a debate. Didn’t archivists reject item-level cataloging decades ago, and for very good reasons? I don’t see that resources are being pulled from item-level cataloging to do digitization – those resources were lost long before digitization came on the scene.

    When it’s formulated more broadly as description vs. digitization, and extended out to processing vs. digitization, yes, I can see that some people are still fighting that fight. Though I have to wonder what it hurts to put a minimal description out there, if the alternative is nothing at all. Because that’s the real choice – something vs. nothing, not something vs. everything. Once you have something, you can always do more, if warranted by whatever criteria you deem compelling. No one is giving us the resources to do everything for everything, and no one ever will. (Vexing as it may seem, that’s not really an archives problem – that’s life.)

    On the description/processing vs. digitization debate: I’m an institutional archivist, I’m constantly asked very specific questions that I have to answer at the drop of a hat, usually by users who will never set foot in my archives, and, even so, digitization certainly wins out for me. Once I’ve got a basic description out there, digitizing as much as I can, and providing full text searching for as much as I can, is most helpful to me and my users, both the ones I can anticipate and the ones I can’t. The overhead for the kind of digitization we do here is low, and I can always go back and do more research or provide more description about what something is if someone needs it. People are always going to want more, but “more stuff online” is certainly what my users care about, regardless of how minimally (or maximally) that stuff is described.

  12. Christine,

    I expected the point of view you represent to be the most commonly expressed by archivists. What interested me about this statement is that when it was presented by the speaker at the conference in Dundee, I saw it via Twitter, and those who were tweeting about it seemed to be embracing the speaker’s perspective. I saw no evidence in the public discussion, as reported on Twitter or blogs, that any archivists were disagreeing with or challenging this statement. Hence my question at the end of the post about there being some perspective on this that we might be missing in the U.S.

    That said, I still think it’s interesting to hear the perspectives of users on this question.


  13. I agree with Christine Di Bella above, long ago archivists figured out that they couldn’t reasonably provide item-level cataloging for most collections. That doesn’t mean that some collections don’t warrant detailed description or even calendars, or that a genealogically useful collection of family trees shouldn’t have at least a list of major surnames, but item-level cataloging for, say, the routine business records of a large mid-century corporation is likely to not be worth the time or effort.
    However, it seems that mass digitization, mostly with folder-level description, as done by the Smithsonian Archives of American Art and the Leo Baeck Institute (two examples I know of), addresses the needs to many (most?) users. For the genealogists, they may or may not be able to keyword search an ancestor’s name, but in Digibaeck they can certainly search for an obscure little town like Unterreisenheim and peruse the community register from home:
    Similarly, if a researcher is interested in the records of art galleries, it’s hard to imagine what could much be better than what AAA is doing, taking existing folder-level EAD finding aids and linking to a series of images that are simply the contents of the folder.–co-records-9936
    Of course, this presupposes sufficient resources for both folder-level description *and* mass digitization, but I suspect this is the route more and more archives will (try to) go, especially if fair use holds up as a copyright strategy.

  14. Good question, and the context reminds us that even though we all view access as important, our definition and our users’ definitions for access may not always match. Item-level description that can support on-demand digitization matches user demand to access; with the appropriate digital repository and processing standards, this can also service digital preservation.

  15. More somewhat random responses from Twitter:

    @badger5 I had this discussion the other day with a co worker in regard to using TEI. We were discussing that digitization is great, but if there’s no encoding for the text then what’s the point in re. to DH.

    @lostinhistory Why not “cowdsource” the digitization and the cataloging can be organized by the archivists?

    @stefarchivist They think they do, but see how they’d complain if you gave them 1 million images & said “no index or catalogue – just browse”

    ‏@jordi78 10m perhaps users themselves could be in charge of the lower level description. They want a full transcription ? Let them do.

  16. Glad it’s stirred up such a debate! Here’s how I covered this in my talk:

    “One thing I do really commend the National Records (of Scotland) for is its online catalogue, which I find to be an incredibly effective tool… The Scottish Archive Network catalogue for local archive holdings and the National Register of Archives for Scotland are equally useful for what they hold, though are incomplete, and I would argue that the continued cataloguing of all materials is much more of a priority than the digitisation of records – we can’t research from documents that we don’t know about. A centralised platform that provides links to these and other externally hosted catalogues would certainly also be a great innovation.

    “Similarly, with indexing, some collections in archives remain virtually impenetrable simply because they are so unwieldy and unindexed – for example the Registers of Deeds – and take a horrendous amount of time to work through, so un-indexed collections equally remain an issue – and again, I would suggest this is equally as important as digitisation.”


  17. Have to agree with Christine and Kevin here– even if we grant that item-level metadata is more useful than digital images of the files, the sheer amount of labor that goes into providing that metadata is such that this isn’t a real choice. Even with born-digital materials, if the records creators are not specifically providing the rich metadata that is being discussed here as useful, extracting that data in a form that is usable by most researchers is *extremely* difficult, often involving a huge amount of manual standardization and massaging. If the materials that researchers want access to are born-analog, with no embedded metadata at all? Oy. Good luck.

    Would it be nice to have individual entries for the kinds of records being referred to here? Sure. Is it *practical* for most archives given limited time and funding? I am skeptical. This might be something that could be crowdsourced, or possibly added to a bit at a time– I am thinking of our process for our University Photo Services negatives, where we are slowly adding images to our digital collection as they are requested by patrons or for exhibits. In terms of a systemic process, though…

  18. Ashamed at being predictable given my specialty, I feel like there’s a format consideration issue missing here. Some type of reformatting needs to occur for most legacy audiovisual materials in order for them to be accessible at all, and there are frequent situations where descriptive cataloging cannot occur until that reformatting has been completed. I think there’s a disconnect in how we think or talk about digitization between digitization for online access via a digital “archive” and digitization that is performed for preservation purposes as with video and audio. In the latter case digitization creates the opportunity — and the need — for item level cataloging. But even before that I feel that, as I’ve written about on our blog, all a/v items need some degree of item level record in order to manage the collection and plan for preservation.

  19. Item-level cataloging?!?!? Really? With my backlog? I’m thrilled with a box inventory for access. Why don’t we just have item-level cataloging for each page of a book?

    Okay, I may sound a *tad* angry. I am getting tired of a few precious items getting item level cataloging because they are deemed “pretty enough” to be digitized. And since they are being digitized in our digital system they need more information then I know about entire collections….but…..I am angry!

  20. Rather regretting my use of the phrase ‘item-level’ in that blog post, which has skewed this debate – it wasn’t what Chris actually said, although I think the phrase ‘detailed cataloguing’ might have been used by one or other of the users represented at the conference. The main tension raised (I think) was between new cataloguing of as-yet-undescribed materials and digitizing well-known and popular series which in the UK are often made available to family historians for a fee. The digitization activities are therefore to some extent self-funded – but in the meantime the materials with no online descriptions remain essentially invisible.

    The archivists in the room were definitely on the side of those invisible archives.

  21. Oh, this is beautiful. I really liked Joshua’s statement because my first thought was that so much of digitization requires item level cataloging, I’m doing some of that lately with videos. But really I’m in line with Christine because I believe you have to know the forest before you can make good digitization priorities and quality metadata for the trees. We’ve been taking that to the extreme by posting groups of accession records by repository in EAD so that we have *some* description online for everything we have;query=accessionrecords;brand=default . We think that may help us attract support for more detailed digitization and description!

  22. If you look a decade out, the production of detailed descriptions is likely to be considered to have been a waste of archival resources.

    In, say, ten years the network bandwidth available to individuals (let alone institutional users) will mean that displaying individual digitised records will be virtually instaneous. I would also hope by this time archives have got their act together and provided sophisticated user annotation systems. The combination of these two technologies will mean that the production of detailed descriptions of records will be produced by outsiders, over time, at their convenience, and to suit their interests and inclinations. At this point there should be no reason for archives to provide any more descriptive metadata than they currently do for physical records – and possibly even less.

    The question then becomes, what should we do now?

    My answer would be: first invest in digitisation. That provides better long term value for your investment, and makes the collection available to those that cannot visit your archive. At the moment you must provide some level of description of the contents – perhaps to the file level, perhaps lower, otherwise users will not be able to enter the digitised collection.

    Second, invest in good annotation systems so that users can add their own gloss to the digitised material, and get the ‘value add’ happening.

    (You will note that underlying this approach is a simple principle: if a particular users want records described in a particular way for their benefit, they can do it themselves. Archives should, however, facilitate this.)

  23. I’m not sure its a case of whether to undertake detailed item-level cataloging or not. For me, as someone working with a national aggregation of archive descriptions, and constantly thinking about how to facilitate discovery, its more about good cataloging. It may be collection-level or lower-level, but sometimes I don’t think that enough thought goes into the choice of words within the description in order to ensure that researchers find key collections for their research. The difference adding a few subject headings makes can really be quite significant, because it focuses their mind on the keywords that might be used to find the description and they may realize that the main description does not contain some key descriptive words that could really help surface the collection.

    I do think that with digitization isome projects do not give priority to the metadata giving context to the digitized item. Images sometimes seem to float around like orphans, detached from their context, nice to look at but maybe of less utility for real research?

  24. From someone involved in digitisation I find that good detailed cataloguing greatly improves the efficiency of digitisation. It also enable us to embed metadata into the files immediately after production.

    On the flip side of this, digitisation can assist the process of detailed cataloguing of collections by retrofitting digitisation workflow tools for the purposes of collecting metadata. Displaying digitised items on the same screen as the metadata fields, copying OCRed text rather than typing etc… can improve the efficiency of the cataloguing process.

    Cataloguing and digitisation should be a symbiotic relationship but resources and processes are often treated separately.

  25. As a historian, I wish I could go to a Web site and find both detailed descriptions of collections and digital images of everything I want to see. Based on what we hear from genealogists, they love what we have digitized and put online that is useful to family research and ask when this or that likewise will be made available to them online free of charge. Non-genealogy researchers also love being able to access digital images online, but the “what” in which they are interested varies depending on the area of research. Some are less likely to want to spend time plowing through digital images and appreciate more detailed finding aids that can get them to the collection that addresses their research needs, but the larger number would probably value having “their” collection digitized.

    As someone who works at an archives, however, I see what many of you experience as well–not enough staff and not enough time with other demands taking key staff away from either type activity. I agree that finding aids are not something that will excite interest from above they way a digital project does. Of course, even if you get the special funds to hire temporary staff and to digitize materials, you don’t get money to support the ongoing technology that will be required to maintain what you’ve done. That, too, isn’t exciting.

  26. As a family historian, collector, and someone who believes in digitalisation of the printed word (for access and preservation) – I believe mass digitalisation of the written word in its various hands will be with us within the next five to ten years.
    For many amateur researchers (such as myself despite several thousand hours in archives) and professionals alike this will make a huge difference to how we relate to both documents and individual archives.
    For the customer the most dramatic change will be the speed, convenience, and depth of search of documents globally – for the archive we will see a blurring of individual archives and focus on training users in accessing the collections with related context rather than visitor search rooms and physical document retrieval and return

  27. We may have a barrel of apples and oranges here. The types of records Chris is talking about above are very standardized documents that lend themselves well to basic and simplistic indexing (registers of deeds and such) and not necessarily the types of complicated photograph or manuscript collections I originally thought of when reading the summary of his conference comments. When processing court records at my previous two jobs, a name index to parties to each case and type of case was a common product of the task, assuming the court hadn’t already done so, and didn’t add much to the workload.

  28. Reading these comments highlighted for me once again that we often use the same terms but mean something different. We have recently been looking at terminology as part of a review of ICA AtoM, and one of the first problems we identified was the use of the word ‘item’. When Australian archivists describe at ‘item’ level, they usually mean the file, volume or folder, not the letters within the file. So here, good quality ‘item’ level description facilitates digitisation, because we have something to link the images too. The National Archives of Australia only digitises material that is described at the file/item level in its online catalogue – facilitating access in two ways.
    On the other hand, digitisation enables indexing – the identification of individuals within a file or volume, often then described in a finding aid, or via a third party site such as Ancestry.
    So when I read a plea by a researcher for more item level description instead of digitisation, I read it as a request for file or volume titles in online catalogues, which will enable the researcher to find the relevant records for their search, rather than a fonds, record group or series description. Once that record can be found, the more detailed description can perhaps be outsourced, via a digitised copy.

  29. Coming a bit late to this discussion. As both a Consulting Archivist (Lone Arranger) and a Professional Genealogist I’m intimately entwined in this issue. I understand the underfunded/no time issue in Archives. I’ve had to choose between processing more thoroughly and digitizing material in order to be able to ‘show’ what I’ve accomplished to a client. I also know the frustration of being a thousand miles away from a collection that may or may not hold a key piece of evidence in a research project and not being able to tell because the Finding Aid and/or Catalog only provided a general description of the material. And the Archives is only open on odd Tuesdays and even Wednesdays (kidding).

    Isn’t an Archives core purpose to provide access to the material it holds? Proper description is the key to that access, regardless of whether it’s in a Finding Aid, an on-line Catalog, or a digitized image. The time it takes to create that proper description is at the heart of this debate (I believe). The reality is the vast majority of Archives are understaffed and underfunded, putting the Archivist(s) between a rock (those running the Institution) and a hard place (the researcher needing access) and making them angry (see Joan G.’s comment).

    As for the pursuit of genealogy there are few subjects in which someone learns the application before the theory but, in most cases, this is what is happens in the family history/genealogy world; the codification of someone’s personal history (entered in a database, usually) comes long before the theoretical aspects of genealogy are explored, learned and applied (such as the Genealogical Proof Standard). However, here’s a real eye opener: in 2012, amateur genealogists (family historians) spent $2.3 BILLION on services. That’s right billion, with a capital B.

    Why is the Archives Industry not actively tapping this Community for support? The power of this group is significantly larger than even they realize. Engaging the more seasoned Genealogical community in a dialog about the serious issues affecting Archives, and therefore the access to material they need for research, has vast potential. They are passionate about the materials they need for their research, often because (speaking from personal experience) there is nothing more moving than touching a document nearly 200 years old that was created by ones own ancestor. The time is ripe to turn the tables and tap into this resource, don’t you think?

  30. Also coming in a bit late, but I hope that I can add some perspective. I do fully believe that ALL OF THE ABOVE (online collection records, online finding aids, and digital content cataloged on the item level) benefit users on different levels. And I also agree that archivists must approach large-scale digitization more creatively. For me, the jury is still out on how effective crowdsourcing may be, but I do know that many of us will be at least trying it in the near future.

    The Archives of American Art, Smithsonian Institution has been digitizing entire archival manuscript collections since 2005. Thus far, we have been able to digitize about 1,200 linear feet totaling 1.7 million digital files. We DO NOT item catalog any of the images for this interface. Rather the EAD finding aid created by the processing archivist serves as the only online descriptive metadata. Thus, we are fully repurposing the EAD metadata that we create as part of the processing workflow. The finding aid also serves as the online navigational tool for the digitized collection. Readers browse the collection by box and folder, just as they would do in the reading room. You may browse our online collections here

    Yes, I am making it sound easier than it is, but this is an approach that has eliminated the most labor intensive step of the archival digitization process – item level metadata. I often refer to our workflow as MLD – Minimal Level Digitization. Although we were digitizing on a large scale prior to the Well Intentioned Practices paper published by OCLC Research in 2010, we were happy to see many of our internal practices and policies further validated by this important report. And while we may have been the first institution to implement this approach, we are certainly not the only one today as more and more institutions find ways to connect and repurpose existing finding aid data to digital content.

    Another report that I would encourage you to read is Supporting the Changing Research Practices of Historians report funded by NEH and published by Ithaka S+R in December 2012. It is available here The study actually supports nearly all of the approaches promoted in the above comments. Yes, “they” do want it all, but they are also reasonable about it.

    Also, to respond directly to Josh’s comments about AV materials. I agree that providing access to these often requires item-level descriptions. AAA is in the midst of a CLIR grant to process collections comprised entirely of AV and collections that contain a large amount of AV. A primary goal of this project is the development of guidelines and methodologies for processing media-rich mixed manuscript collections according to an archival approach, rather than the item-level approach to media typically taken by archives – or when item level identification and description simply may be feasible for the present. At this time, we are drafting guidelines for processing AV at minimal, intermediate, and full levels.

    Barbara Aikens
    Chief, Collections Processing
    Archives of American Art, Smithsonian Institution

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.