Book Review: Everything is Miscellaneous, by David Weinberger

Everything is Miscellaneous, by David Weinberger

Everything is Miscellaneous, by David Weinberger

In Everything is Miscellaneous, Dave Weinberger argues that classifications that we have imposed on most everything from the alphabet to the encyclopedia, planets, books, and knowledge ultimately represent our own beliefs and priorities. As time changes, we see how our own thinking at that time inclined us to organize the information that way.

In reality, things in the world don’t have such clear-cut categories and ordered absolutes. The world around us is ultimately miscellaneous, meaning, without clear division and order. Each thing has characteristics that overlap with other things, so that a classification that makes sense to one person doesn’t make sense to another.

Even if we meticulously craft a careful organization of the content, our organization might not ring true to another’s information needs. For example, organizing content on a map to include topography, water systems, roads, and trails might meet one person’s needs, but the omission of businesses, schools, boundaries, and emergency shelters might make the map useless to others.

Rather than classifying things in absolute ways, or deciding what to include and exclude in systems, we should strive to add as much metadata to things as we can imagine — and allow users to do the same. Applying metadata to everything allows us to manipulate the information in ways that make sense to each of us.

As another example, tagging photos in systems like Flickr allows you to retrieve virtually any set of information based on your queries. For example, you can search for all panorama beach photos in California containing sunrises and see a list of results. But if a group of librarians had forced the photos into rigid collections, their categorizations probably wouldn’t have accommodated this specific category.

Weinberger argues that metadata is really the key to solving the problem of order and classification. Adding metadata should be one of the top priorities of help authors, in my opinion.  Who hasn’t wrung their hands in agony over where to put a help topic. We meticulously try to arrange topics into neat topical hierarchies, but these hierarchies often fail to communities of users. Trying to arrange content in the “right order,” usually based on tasks or topics, inevitably frustrates users. They end up playing guessing games about how things might be named and categorized. After browsing for a few seconds, they resort to hopeless searches, usually trying to guess the right terms.

Even if the order makes logical sense to many users, people don’t always think in topical ways. Maybe users are browsing to discover more advanced techniques, or are looking for troubleshooting issues, or are trying to locate specifications. If they could leverage metadata to arrange the information they need, they’re more likely to find the right topics.

Recently I was working on a project documenting a simple online calendar. In this calendar, administrators can reserve blocks of time for different groups, so I put this topic under a section called “Managing Locations and Resources.” Where would users — not administrators — look if they’re trying to figure out how to submit requests to reserve calendar times?  And where would calendar schedulers look if they’re trying to view resource availability? Would they look in the same section where other administrator topics are grouped? Or in their own grouping of topics that relates to their own rights and roles?

Surely every help author who has documented software that contains at least 75 topics or more has wrestled with this conundrum. This is the threshold in which figuring out the right categories or folders for information poses challenges. It’s where you realize the content could be organized in a number of different paths based on one’s needs, priorities, and perspectives. To users unfamiliar with an application’s rights and roles, unaware of its terms and workflows, figuring out where a topic is located within a large system of folders is a guessing game.

Amazingly, Weinberger doesn’t use help systems as examples of miscellaneous. His domain is instead the Internet itself. But I have never read a more relevant book for technical communicators than Weinberger’s Everything Is Miscellaneous. Weinberger provides a strategy that, if followed, would largely solve the problem of findability with help. In a world of ever-increasing information, adding metadata to push and pull and arrange the information in user-centric ways provides the best strategy for attempting to organize content. Ultimately, Weinberger says to give up on organizing the content in one specific way and instead allow users to group it in ways that make sense to them.

The challenge, as always, lies in implementation. Exactly how do you add metadata to your help topics? What kind of metadata do you include? And how do allow the user to arrange or call the topics based on the metadata they want to sort by?

Weinberger is a masterful writer with a strong knowledge of history, philosophy, Internet, politics, and more. He weaves together stories from different disciplines in a mesmerizing way while always tying in his overall thesis — that everything is ultimately miscellaneous (without absolute order and classification) — to each chapter.

In the second half of the book, Weinberger focuses on the implications of not having a centralized, top-down hierarchy of knowledge. Some of his philosophical interests surface more visibly here, but overall the book is extremely relevant to the field of technical communication. For anyone who has ever struggled to organize help content, this book provides solid answers and strategies.

[amazon-product align="none"]B0029PSOOS[/amazon-product]

follow us in feedly

Madcap FlareAdobe Robohelp

This entry was posted in book reviews, findability, general on by .

By Tom Johnson

I'm a technical writer working for a gamification company called Badgeville in the Silicon Valley area in California. I'm primarily interested in topics related to technical writing, such as visual communication (video tutorials, illustrations), findability (organization, information architecture), content development (DITA, testing), API documentation (code examples, programming), web publishing (web platforms, Web design) -- pretty much everything related to technical writing. If you're trying to keep up to date about the field of technical communication, subscribe to my blog either by RSS or by email. To learn more about me, see my About page. You can also contact me if you have questions.

12 thoughts on “Book Review: Everything is Miscellaneous, by David Weinberger

  1. B Noz Urbina

    I LOVE this post.

    Metadata and organisational systems are a particular are of fascination for me. I’ve had many experiences where authors have tried to create systems and share them with colleagues to endless frustration. And your pointing out that metadata is of such primary importance is dead on. One could guffaw in modern times at those who are still working in a monolithic, print-based paradigm, but without metadata, we’re making even searchable, electronic deliverables more static and inflexible than we need to.

    There’s much talk of the threat of ‘social media’ to Google search. The power of so much socially added metadata means that people will go to people as filters rather than have Google do brute force analysis on millions of irrelevant potential web pages.

    Search is not always as powerful or appropriate as facetting or facetted search (navigation and search within certain metadata parameters).

    IBM’s Sophie McMonagle (http://bit.ly/gruFBo) is discussing at the Congility 2011 Conf how their team organised a portal delivering four million web pages and one million DITA topics. 5 million objects…? The thought gives me a facial twitch.

  2. Marcia Johnston

    A fascinating topic indeed. Sounds like Weinberger’s treatment of it is worth reading. Thanks for the review, Tom.

    You say, “Amazingly, Weinberger doesn’t use help systems as examples of miscellaneous. His domain is instead the Internet itself.” It amazes me, too, that so many discussions of content strategy and information architecture are Internet-centric. Here’s to findability for the rest of us!

  3. Eddie VanArsdall

    Tom, you are downright influential. This book has been on my to-do list for some time, and after reading your review, I downloaded it to my Kindle. Yes, I’m a Kindle nerd.

    Like you, I ponder over the conundrum of making help (and content in general) searchable. You and I have discussed the pros and cons of what systems such as Flare offer. The features have great potential but often fall flat. We need our tools to function more like a mini-CMS. Why not provide an automated mechanism for adding metadata at the topic level, like DITA does? Why not implement a GUI-based feature for faceted search?

    Caveat: I still believe that taxonomy has its place in certain contexts. A well-constructed thesaurus can be quite useful in guiding users to specific terms and concepts, especially in specialized fields. Building ontologies in languages such as OWL and RDF is especially relevant to complex fields such as genetics. Dismissing these methods as print paradigms is simply wrong. They can be quite effective on the web.

    As I see it, a combination of search mechanisms works best: metadata, controlled vocabularies, facets, etc. Tagging alone can be a real nightmare, especially when users are searching for meaning.

    1. B Noz Urbina

      Hi Eddie, you seem to be making some distinction between a taxonomy and metadata. Interesting. Although you can apply metadata freely and allow users to sort it out, I feel that if metadata is really going to be effective you’ll need some sort of taxonomy to provide rails. User additions are extremely vauable, but they can be additions pinned to a framework, rather than waiting for a whole folksonomy to develop before you know if that folksonomy is even effective…

      1. Eddie VanArsdall

        Hi Noz,

        Metadata and taxonomy are two distinctly different approaches to information retrieval, but they can work well together. My point was that I think it’s short-sighted to dismiss taxonomy as the work of “a bunch of librarians.” In fields where semantic relationships are key to true understanding (I used genetics as an example), metadata is a complement to, but not a substitute for, a robust taxonomy.

        While user tagging is effective for establishing trends and giving users a voice, it eventually becomes unwieldy and imprecise. Someone eventually needs to play custodian. That involves learning from the trends, cleaning up the tags, and integrating those tags into more structured metadata.

        So what I’m advocating is giving users multiple methods of retrieval, rather than simply declaring that metadata is all they need. That also includes a search box and a search algorithm.

  4. Jonatan Lundin

    Building taxonomies and categorizing content is a fundamental task, especially in our domain (technical communication). But maybe more important is to know what content to write. What type of information are end users searching for?

    Just ad-hoc writing *something* and then, when your done writing, develop a user-centred taxonomy where your *something* content is categorized wont help users. Maybe the content they are looking for doesn’t exist or even worse, a topic having a metadata that the user is looking for is treating something completely different (topic content and metadata classification mismatch).

    Put it the other way around; first develop a taxonomy where the taxonomy mirrors what type of content end users need (this is actually what you do when conducting an information analysis etc). Then you can start writing and as you complete topics they can be categorized.

    The SeSAM framework is a “ready-made-default-taxonomy-for-technical-communicators” where the main facet is the search situations. One of my hypotheses in my research, and the foundation for SeSAM, is that it becomes easier for end users to find if they understand (and accepts) the way you are thinking when developing content. Meaning that the user becomes aware of the classification and organization principles.

    Tech communicators are putting a lot of effort in thinking about organization and categorization, but how much of the thinking is communicated to end users? Do they understand how you’re thinking? My believe is that one key to findability is to communicate the strategy. Compare the findability in a grocery store where you know how groceries are organized compared to another grocery store where you’re not sure about how they have organized things. You know (maybe) they have it but where is it located?

    1. Tom Johnson

      it becomes easier for end users to find if they understand (and accepts) the way you are thinking when developing content

      It’s fine to present the information this way, but in a digital world, where you can combine and rearrange and sort content in myriad ways, why not provide a variety of organizations for users to choose from? Isn’t it a bit impractical and writer-centric to expect the user to learn and understand our way of thinking only?

      1. Eddie VanArsdall

        Yes, and I think the taxonomy community realizes that, too. In the more modern books and teaching, they’re advocating a variety of search methods, including tags created by users.

        1. Jonatan Lundin

          For me there are two sides of the coin. One is communicating what type of content the user can expect to find in your deliverable and the other is how the content is organized. Your deliverable can communicate both of them or just one, or none. So I do not see a contradiction here; you can provide a variety of organization schemes for the same type content. Users are different and have different starting points.

          But I still argue that findability will increase if the user has understood what type of content your deliverable contains and how it is organized. The question is how we make the user understand.

          Providing several organizational schemes may infact mean that users gets a better chance to understand what type of content your deliverable contains, since a content snippet has several facets. Another way is to explicitly communicate your strategy “This manual contains information of type X, Y and Z”. The content shall of course, by itself, as much as possible communicate its meaning. But I see many cases where it is impossible to understand what type of content is hidden behind a certain title in a manual. What is the writer trying to say? A lot of metadata will indeed help the user to understand what it is all about.

          We as tech communicators shall of course not force users to learn our way of thinking. And if the user re-arranges the content or ads new metadata, they will be in charge and possess the knowledge about how the content is organized.

          An interesting hypotheses regarding findability is standardizing. Let’s imagine that we could come up with an architectural scheme that would be suitable for all types of manuals for technical products (incl software). Meaning the type of information to include and how it is organized. A user that has learnt the scheme will have no problems in finding information in a manual for a completely different product? Is it possible to come up with such a scheme for technical manuals? I think so.

          The idea of standardizing as a fundamental aspect of usability is well known (check for example Donald Norman) and used for example in cockpits, car interiors and even daily newspapers follows the same kind of scheme (almost all newspapers have chapters for international news, domestic news, sports, culture, etc – but the way this content is organized may differ from paper to paper).

  5. Pingback: Topics should merit their metadata

  6. Pingback: What Does “Every Page Is Page One” and “Include It All, Filter It Afterward” Mean? | I'd Rather Be Writing

  7. Pingback: What Does “Every Page Is Page One” and “Include It All, Filter It Afterward” Mean? » eHow TO...

Comments are closed.