The day is April 4. I extract the half-and-half cream carton from the work mini-fridge and prepare to unscrew the cap when I notice the expiration date printed at the top. “MA 14.”
Does “MA 14” mean March 14 or May 14?
I stop. Is that March 14 or May 14? Both months begin with the letters “MA.” Both dates seem plausible when I consider that it’s early April. This half-and-half cream might have gone bad two weeks ago or it might have two weeks of quality left in it. Which is it? Is this going to give me a tummy ache or not?
Fortunately, I have not exhausted the backup plan. I have used only a fraction of my power. I unscrew the cap and hold the carton to my nose. Sniff. Seems okay. The half-and-half plops into my coffee.
Afterwards, I ask my co-workers in the office what they think “MA” means. My Canadian co-worker is confident that it’s May.
Quickly I get to wondering about what a clearer abbreviation might have been. “MY,” perhaps? March doesn’t have a “Y” in it. Maybe a number scheme would have worked better – the carton could say the number of the month (“05”) instead of its name. But without context, that digit is simply data, not information. It’s too vague to mean anything. Even if the user knew that number signified a month, it is still probably easier to more quickly recognize a month by its letters than by mentally retrieving the other half of the key-value pair “fifth month.”
Then I remember that being a product of Canada, an officially bilingual country, the month abbreviations surely must have been carefully chosen to fit both English and French spellings. After all, it would have been even more confusing had the carton contained two month abbreviations, one for each language.
Curious, I later poked Google for anything that had been written about Canadian month abbreviations to see if I was right about the labels needing to accommodate both official languages. Yes, turns out that is indeed the case: the two-letter abbreviations were designed to contain only letters that show up in both languages’ spellings of each month.
JA
January
janvier
FE
February
février
MR
March
mars
AL
April
avril
MA
May
mai
JN
June
juin
JL
July
juillet
AU
August
août
SE
September
septembre
OC
October
octobre
NO
November
novembre
DE
December
décembre
Two-letter Canadian month abbreviations
“Canadian Month Name Abbreviations,” a minimally styled HTML page from 2014, explores the topic in a satisfying level of detail. The author ruminates on other possible abbreviations for each month and realizes, as I begrudgingly did, that “MA” was literally the only possible option for May given the constraints on the naming conventions for this information.
The author reverse-engineers the information architecture in detail:
The abbreviation starts with the first letter of the English or French name. (By chance, English and French always agree on the first letter. Not true with, for example, English and Spanish: January/enero.)
The second letter has to be in both the English name and the French name. Consider the month April/avril. It can’t abbreviate as “AP”, because there’s no “P” in “avril”, and it can’t abbreviate as “AV” because there’s no “V” in “April”. But both “April” and “avril” have “L” in common, so they can (and do) abbreviate as “AL”. (The other possible solutions: “AI” and “AR”.)
As a user of the half-and-half, “sniffing” for information about the expiration date of the product, this particular labelling scheme failed me. I was unable to discern meaning from the two letters I was provided. And yet, my Canadian colleague could immediately tell what that abbreviation meant. The fact that the two-letter code must accommodate both English and French was a “a-ha” moment to me, a recent immigrant from the United States, but it was obvious to her. A mere abbreviation on a carton reminded me that I was a foreigner.
However, given the bilingual constraints of this labelling scheme, I think the information architects at the Canadian Food Inspection Agency did the best they could given what they had to work with. I put this new information into the back of my mind, and the half-and-half carton into the back of the office mini-fridge.
Soon after this pronouncement, I noticed an interesting phenomenon begin to occur in the comments sections of YouTube videos. Users started making their own “dislike buttons.”
Other users simply left one word comments that said… you guessed it.
As an information architect, I’m fascinated by this kind of push-and-pull dynamic between the top-down directives of the platform and the bottom-up behavior of the users. Clearly the users of YouTube still want a way to publicly express their dislike of a video, as evidenced by the comments section workarounds they have come up with.
In a video accompanying the blog post, Matt Koval, Creator Liaison for YouTube, describes this change as mitigating a “whole other use” for disliking a video, which is to harass the creator of the video through something YouTube calls “dislike attacks.” At what point does a sincere dislike become an “attack” dislike? Your guess is as good as mine.
Typing out the word “dislike” is certainly a “whole other use” for a comments section. Will that be the next user forum to get gutted? I have to say, if curtailing harassment was the goal of this change, it’s utterly baffling why the comments section wouldn’t have been the first feature to be axed. I genuinely wonder if YouTube will eventually turn comments into some sort of one-way function where users can leave a comment that only the creator will be able to read, much like those suggestion boxes where one drops a folded up slip of paper into what feels like a void.
Pictured: The future dynamic of the YouTube comments section?
I also found it amusing that one of the reasons Koval names for the dislike count being “a big problem” is because “half of YouTube’s mission is to give everyone a voice.” How is removing one of the key mechanisms for interaction between users anything but the precise opposite of that mission?
Besides, creators can still create their own commentary videos about why they dislike something. Surely several minutes worth of spoken commentary (plus the view and like counts of said commentary) has at least as much potential to “harass” as a thumbs down?
Indeed, there’s been no shortage of commentary from prolific, established YouTube creators about why this move makes little sense and worsens the user experience of the platform. One of the best I’ve seen is from Marques Brownlee:
One of the key points Brownlee makes is that the dislike count is not merely that—it is also one of two necessary halves of a like/dislike ratio that provides “the immediately glanceable piece of information that I can look at when I arrive at a video to know if it’s going to be worth my time.” Neutering dislikes also neuters likes.
If I see that a video has 3,000,000+ views and 27,000 likes, what does that mean? Is that… good? Is that a particularly well-received video? It’s widely acknowledged that the likes-to-views ratio on YouTube has always been low (I’ve seen several sources put it at 4:100 on average). Most users only watch the video and leave as soon as it’s done, even if it’s a video they enjoyed. I can’t assume that a like count that’s far below a view count is necessarily “bad.” The entire like/dislike mechanic is effectively meaningless to me now, save for whatever unseen, mysterious influence my choice might have on “the algorithm.”
Compare the image on the left to the image on the right. Which UI provides the user with the most helpful information in the shortest amount of time?
Which UI provides the user with the most helpful information in the shortest amount of time?
If a user clicks on, say, a how-to video (or perhaps a video about COVID-19) and immediately sees a disproportionate number of dislikes, that’s an immediate indicator that the video is suspect—and not necessarily because the information is malicious or the creator is untrustworthy. It might simply be that the information is out of date, for example. But now the user will need to either sift through YouTube’s notorious comments section for a clue as to the general reception of the video (which seems less than ideal if the goal is to decrease harassment) or they will need to waste time watching more of the video than they otherwise would have because they need to figure out for themselves whether it’s worthwhile.
Hm, I think I may have hit on something. One of the side effects of this change is that a greater number of YouTube videos will attract lengthier watch times. That’s fantastic from the perspective of YouTube, their advertisers, and their shareholders, who want to be assured by perpetually climbing eyes-on-the-screen metrics. But from the perspective of YouTube users? Not so much. Think of it this way: if I wanted to maximize the time users spend on a piece of written content trying to find the information they’re looking for, I could remove all the headers and equalize the formatting. That would certainly boost time-on-page, but would it help usability? What would that say about my priorities as an information architect?
Information architecture is a balancing act between the business needs of stakeholders and the desires of users. Unfortunately, the decision of YouTube to remove the public dislike count has to be one of the most lopsided moves I’ve seen from a major platform in a long time. It is entirely in favor of the platform at the expense of the creators and users. I say this confidently because I quite literally have not seen a single actual YouTube creator speak out in favor of this decision. To quote Jawed Karim, YouTube co-founder and star of the first video ever uploaded to the platform:
Calling the removal of dislikes a good thing for creators cannot be done without conflict by someone holding the title of “YouTube’s Creator Liaison”. We know this because there exists not a single YouTube Creator who thinks removing dislikes is a good idea—for YouTube or for Creators.
Why would YouTube make this universally disliked change? There is a reason, but it’s not a good one, and not one that will be publicly disclosed. Instead, there will be references to various studies. Studies that apparently contradict the common sense of every YouTuber.
The ability to easily and quickly identify bad content is an essential feature of a user-generated content platform. Why? Because not all user-generated content is good. It can’t be. In fact, most of it is not good. And that’s OK. The idea was never that all content is good. The idea WAS, however, that among the flood of content, there are great creations waiting to be exposed. And for that to happen, the stuff that’s not great has to fall by the side as quickly as possible.
The process works, and there’s a name for it: the wisdom of the crowds. The process breaks when the platform interferes with it. Then, the platform invariably declines. Does YouTube want to become a place where everything is mediocre? Because nothing can be great if nothing is bad.
In business, there’s only one thing more important than “Make it better.” And that’s “Don’t fuck it up.”
I’ll take a guess as to the reason: I think this change is to appease the “large” creators, despite YouTube’s insistence to the opposite. No big-name movie, music, or game studio enjoys seeing their videos added ignominiously to a list of most-disliked YouTube videos. I’m going to go out on a limb and guess that they are probably less inclined to funnel marketing budgets towards platforms that have humiliated them. I’m not the first to posit this, although some other creators lend this idea less credence than I do, arguing that even the most thoroughly embarrassed brands see all press as good press. My retort: even if brands are willing to tolerate the short-term sting of shame, other entities might be less so. Some of those entities might even be closer to home than you realize.
Fortunately, there does exist a way to return the YouTube dislike counter: returnyoutubedislike.com. I’ve installed the Chrome extension. I would encourage any of my fellow daily YouTube viewers who are annoyed by this change to do the same. Judging from the resurfaced dislike counter on YouTube’s pronouncement video (pictured above), I can tell there are at least several hundreds of thousands of you.
To those out there who agree with YouTube’s decision, I ask the following questions in full sincerity:
Would you support removing the ability for users to give one-star reviews on sites such as Etsy, Yelp, or Amazon in order to protect the mental health of merchants?
Would you support the criminalization of the Better Business Bureau and similar consumer protection nonprofits in order to protect businesses from possible targeted harassment?
Would you support the abolition of public election result data in order to protect defeated candidates from embarrassment?
For the sake of usability, transparency, and democracy, I hope your answers to those questions are a big, fat 👎.
Towards the end of July, I received an email from the newly created WIAD Organizer Alumni newsletter. I learned that the Global Board is looking for input into a rebranded name and tagline, since both the event and the nonprofit entity behind it exist as separate but identically named entities.
The email linked to a Google Form survey. I thought I would give my thoughts behind my answers, in case it can add to the discussion.
Brand Name
“World Information Architecture” on its own seems too barren to me. It could be another subdiscipline of information architecture (e.g. “pervasive information architecture,” “product information architecture”) rather than an organization.
“World Information Architecture Association” seems okay, other than the potential for it to be pronounced as “double-u i double-a.” Too many doubles. “Double-u i a a” sounds like a new type of battery or a meeting for alcoholics anonymous participants who are shouldered with something that sounds like a “double DUI.”
“Community” seems like something too broad to claim dominion over. “Movement” sounds political.
“World Information Architecture Organization” is the clear winner for me. A big indicator that this is the right choice is that the explanatory text within the survey already makes reference to “our organization.” It comes naturally!
Tagline
Out of these choices, “Making information clear” is the closest to a succinct explanation of what it is information architects really do. (I think “making information understandable” would be even more accurate but it doesn’t make for a strong tagline.) We’re not community organizers first and foremost, even if some of the work we do does end up connecting communities.
Good luck and thanks to the Global Board for their ongoing efforts toward putting on World IA Day 2022!
Last month I visited my family in America for the first time in quite a while. I didn’t bring any books with me so that I might instead read something from my stateside shelf. Richard Saul Wurman’s Information Architects immediately stood out. I hadn’t actually read it cover-to-cover until then. With many of its 232 pages devoted to large photographs, drawings, and other graphics, I managed to get through Information Architects in only a couple of days. I wrote the following reading notes and thoughts soon after:
I’ll definitely need to add the below video to my “Watch Later” list. Information Architects opens with a heartfelt dedication to Muriel Cooper, a designer at MIT whose “Information Landscapes” presentation at the TED5 Conference “changed forever the visual paradigm of information for all who saw the presentation,” according to Wurman. It’s incredible that the very same presentation that so heavily influenced Wurman – and thus, our whole field – is readily available.
The front cover of the book prominently features a three-part definition of “information architect” that I’ve seen quoted a million times, but the back inside flap offers a concise summation of “information architecture” that I’m surprised I haven’t seen referenced more often. “The design of understanding.” That’s it! Can you come up with a better, more straightforward description of IA?
Somehow this detail was lost on me up until now, but most of the book is a collaborative effort between Richard Saul Wurman, his editor Peter Bradford, and a couple dozen other designers. Wurman reached out to a number of designers and asked for “extended and I mean extended captions” of specific examples of their work, and the pieces were compiled and designed into this book.
Each section of the book highlights a designer or team of designers and walks us through how they were faced with a topic that needed clarification and how they went about creatively clarifying that topic with their design skills. These design challenges ranged from city maps to museum exhibits to product brochures to CD-ROM encyclopedias. Only one of the examples had to do with organizing content on a web site. The case study pattern of organization suits the book well.
The diversity of domains featured in the book leads one to believe that information architects are all around us (whether the IAs themselves realize it or not), yet the foreword curiously describes a “relatively small world of information architects.”
Instead of going step-by-step into how one can practice information architecture, the book serves as a collection of examples of how creative, thoughtful structuring of information led to better understanding. As a source of inspiration, Information Architects is good; as a textbook, not so much.
The book concludes with a dramatic example of the clear value of information architecture: Alexander Tsiaras explains how a three-dimensional volume rendering of the brain enabled by CT scanning allowed surgeons to perform a cranial operation on a child. (“Surgeons say that before volume rendering, surgery was like walking into a dark room and feeling your way around the furniture. Now it’s like walking into a room, turning on the lights, and seeing exactly where you’re going.” Page 227). This new way of visualizing data let radiologists, whom Tsiaras likens to “hermetic cubists and abstract expressionists” speak in a language familiar to surgeons, whom Tsiaras compares to sculptors.
Some of the examples are very much a product of their time. This book was released in 1997 and it shows. Peter Bradford’s attempt at a “curriculum dictionary” that attempted to completely reorganize a dictionary in such a way that readers could identify and traverse related word groups seemed like an unavoidably messy attempt at something that only Wikipedia could eventually accomplish, which is to say a bottom-up information architecture.
The job of an encyclopedia is to explain, isn’t it? Well, how can that be done best? With vertically deep, exhaustively detailed explanations like the Encyclopedia Brittanica, or with horizontally broad, relational explanations like the Curriculum Dictionary? Maybe both? Yes, I think so, too. But, how does one build such a mass of linked knowledge?
Page 73
We made context pictures for easily pictured word groups like Bodies of Water, tables and typographic diagrams for less easily pictured groups like Poetry. Very quickly, our representations multiplied and grew to unwieldy size. They began to crowd the alphabetic section, making it jumpy and difficult to use. To accommodate them, we tacked on a group of pages after the alphabetic section and called it our topic section. Well, not so easy, Sneezy. Topically arranged reference is neither familiar nor encouraged by American publishers. In fact, splitting the dictionary into two sections was to become our most provocative change. But how could we deny the logic?
Page 71. This particular project was ahead of its time.
The 2021 Society for Technical Communication Summit is coming up soon, and that means the conference proceedings have been released. It’s cool to see that my paper, co-authored with Peihong Zhu, is the first article in the bunch. See the standalone PDF below, and attend the virtual Summit June 7-9, 2021 to watch our full presentation.
Josh Anderson 