UK government changes website guidelines due to buggy screen readers
insidegovuk.blog.gov.ukI think a lot of commenters in this thread are missing an important point about accessibility. It does not just mean "works with screen readers". It means easily understood by anyone. A lot of people confuse the terms e.g. and i.e. so it is normally better to say "For example" or "That is" because these are natural phrases that will be clearly understood by a greater number of people.
I try to constantly improve my writing style, where improve normally means "simplify". I am personally guilty of overusing these abbreviations through force of habit, but I edit them out when I can. For anywhere with a formal house style, adopting this seems to make sense, even for things like scientific papers. Excessive use of jargon is a common accessibility problem and in most cases, there's no good reason for it other than dogma.
I am also a big fan of the old Borland "no nonsense licence" for similar reasons. For anyone that hasn't read it, here's a link:
http://www.osnews.com/story/22342/Borland_in_the_1980s_Treat...
This just illustrates that even something requiring the precision of a legal document can be written in plain, approachable English. There really is no excuse.
Many documents and websites produced by government in the UK have the "Crystal Mark for Plain English". Details at [1].
I don't notice the difference, but non-natives have occasionally said they've been surprised when certain information is clear, such as letters from a bank.
[1] http://www.plainenglish.co.uk/services/crystal-mark.html
You probably would have noticed the difference in the 80s and 90s when banks and others still wrote in tortuous, horrifically formal, English. Why take two sentences when a page and a half would do?
Usually the only places you still regularly saw "inst", notwithstanding and heretofore. If you wanted to understand at first reading being a contract lawyer helped.
The Plain English Campaign took out most of the low hanging fruit years back.
I love what UK Gov is doing, they actually sound like the kind of business that I'd love to work with (I'm totally surprised because it's a Government) and I'm learning a lot just with these articles.
I'd also love to see other governments to take a similar approach, but that seems highly unlikely for most countries including Spain (where I am from).
NHS is also very good in my opinion. http://alpha.nhs.uk comes in at 100KB with fonts
We’re hiring back-end developers and web-ops currently, maybe you should apply?
Thanks, it looks great and I'd consider it in 8-10 months, but I am committed to doing some contracting/projects right now.
How is this relevant to the conversation?
Because they are working on UK Government projects, check the profile :)
It has worked because they've done what goes against normal government behaviour and that is, "keep it simple". They are also learning to share resources so that multiple government sites can share postal and payment engines - something which is so obvious but so lacking!
Aside from the quirky interpreations of "eg" and "ie" that triggered this decision, the abbreviations "e.g." and "i.e." are actually problematic, especially when dealing with non-native speakers (or even simply people who had no exposure to Latin).
"e.g." means "exempli gratia" -- I am fairly confident the majority of readers in an international audience (or maybe even a national one) will not actually know this and come up with their own personal backronym (like "ergo" or "example given").
"i.e." means "id est" (literally "that is") -- I know for a fact that even people with some exposure to Latin get this one wrong and instead think it means something like "in exemplum", which actually leads to people using it incorrectly when they should actually use "e.g.". I know this is what I did as a teenager in Germany (while learning Latin in school) and I know that this is what many of my German colleagues tend to do, simply because it seems like an obvious equivalent of the German "zum Beispiel" ("for example").
In the context of a website that explicitly tries to use simple English when possible, I think it is perfectly valid to preempt this confusion and not use phrases which are not actually English and can trivially be substituted with unabbreviated English equivalents in prose (whereas even the unabbreviated Latin would likely not help the reader).
EDIT: I'm not saying you shouldn't use "e.g." or "i.e." ever. I personally use them all the time. But it's absolutely consistent to avoid them when you're trying to use simple language to be understood by a broad audience with varying levels of comprehension.
I don't see what knowing these abbreviations has to do with knowing Latin. They are, to all intents and purposes, English words used in English writing, and people who have had the opportunity to develop a good English vocabulary in other respects will have learned these words too.
(Note, I grew up in England. I have begun to suspect that these abbreviations are used a little less in the US. But there again, if that is the case then the problem still lies with people's unfamiliarity due to not much usage — or perhaps with screen readers' unfamiliarity due to US-centric development — not with Latin.)
(Also, wouldn't the German equivalent of "i.e." be "d.h." which I'm sure I've seen for "das heißt"?)
You are right about "d.h." being the German equivalent of "i.e." but I've met a lot of Germans who pronounced it "daher" ("therefore"). I guess that proves a point about avoiding obscure abbreviations when someone else needs to understand what you wrote.
> come up with their own personal backronym (like "ergo" or "example given").
I'm not a native English speaker, and I've never studied Latin, but I do have access to Wikipedia, Google, and dictionaries, and a policy of not using words I don't understand. If you come across something you don't understand, there really is no excuse for making something up instead of looking it up. It wouldn't be difficult to avoid horrors like "per say" or "segway".
You severely overestimate the eagerness of the vast majority of people in the vast majority of situations.
Also, if when learning English you didn't use any words you didn't perfectly understand, you are most certainly in the minority.
A comment on another site said that the author had heard someone claim ‘eg’ was short for ‘egsample’.
> And while ‘e.g.’ gets read correctly by screen readers, there are better, clearer ways of introducing examples for all users.
Strongly disagree. "e.g." has its place; it is often clearer than any alternative. Likewise "i.e.". "eg" is unpleasantly confusing even as an ordinary reader and should never have been allowed, but to remove "e.g." too is to throw the baby out with the bathwater.
Could you explain why it is better than, say, "say"?
(neat, huh?)
Because there are cases where that's a weird thing to say, say, this exact sentence.
They carry different connotations; "say" suggests a single example that is not necessarily representative but rather being used to make it easier to express the following part, whereas "e.g." can mean a list, and puts the emphasis on the examples themselves. One isn't better or worse than the other; rather both have their use cases.
Because e.g. is formal and "say" is informal.
They could put the e.g. inside a dfn tag or teach the screen readers how to read it.
> They could put the e.g. inside a dfn tag or teach the screen readers how to read it.
I'm afraid not. AFAIK, screen readers don't apply any special processing for DFN tags. Speaking as someone who's using a screen reader to read this thread and proofread this comment, I'm not sure I'd want them to either, and I hadn't even heard of them until today.
Also, screen readers are able to speak through several, possibly hundreds, of different speech engines, so you would have to teach each of those engines to deal with these corner cases separately. Given that many of them were developed ten plus years ago and are now no longer under active development despite continued widespread usage, that's not really a realistic goal.
N.B. AFAIK is spoken by my screen reader/speech engine combo as "uh-fake". As a blind software developer, I have much bigger problems to worry about than memorising how my software speaks abbreviations. I of course don't speak for everyone though and clearer communication is always an excellent goal.
In what cases would "e.g." be better than "for example", aside from when you're desperately short on space?
Most of the cases I see "e.g." and "i.e." -- with or without the dots -- it means "I'm putting zero effort into putting these phrases together into a clear sentence."
People will use buggy software, and I think some of the things it might say would be quite amusing:
"If you have committed any crimes egg tax fraud, insurance fraud, theft"
Egg tax fraud anyone?
Writing 'eg' instead of 'e.g.' or 'ie' instead of 'i.e.' is just dumb, though. So is confusing the two.
The Economist recommends "eg" (no full-stops) in its style-guide[1]. As does the Guardian/Observer[2], and the University of Cambridge[3].
There are counter-examples, such as AP and NYT, but this usage - without full-stops - is not uncommon or abnormal.
[1] http://www.economist.com/style-guide/abbreviations
[2] https://www.theguardian.com/guardian-observer-style-guide-e
[3] https://www.cam.ac.uk/brand-resources/guidelines/editorial-s...
1. Not everyone is a fluent English reader. If you are providing government services, you have to provide them to "dumb" people as well. "Dumb" in this case includes a lot of smart people who were not raised with English as their first language.
2. There is no official version of the English language. Pronunciations, spellings and style guides change over time.
Disclaimer: This post is subject to Muphry's Law: https://en.wikipedia.org/wiki/Muphry%27s_law
> Not everyone is a fluent English reader.
I.e. & e.g. aren't English, but rather Latin. Any educated person in the Western world should understand them, no?
The comments section of the blog post is a perfect example of bike-shedding.
A lot of developers aren't familiar with ADA compliance, it could become something we're (edit all US based developers) all talking about in the near future.
http://www.technologylawsource.com/2015/06/articles/informat...
This is only a requirement for government entities and non-profits at the moment. For those who didn't read the article, in April of 2016 the Department of Justice, the organization responsible for the Americans with Disabilities Act, has proposed an amendment, which will "require public entities and public accommodations that provide products or services to the public through websites on the Internet to make their sites accessible to and usable by individuals with disabilities.". Immediately after, the article says that they were careful not to suggest that currently inaccessible websites are exempt.
So if you have a website currently up and you're providing a product or service, similar to having a wheelchair accessible entrance, you may be required to provide a more accessible website, or an accessible alternative.
For those of you who aren't familiar with website accessibility, you might want to brush up on the ADA guidelines. https://www.ada.gov/pcatoolkit/chap5toolkit.htm
I tend to test run my websites through Lynx to make sure it works fine without JavaScript, try to navigate with only a keyboard in Chrome and Firefox, and try it out with a cheap or free screenreader like Chromevox.
FFS can we stop spending billions on designing around screen readers and fucking spend billions on good screen readers.
Even if that happened it would be years before people upgraded. People using assistive technology tend to be more resistant to changing their setup than one would expect as it can be difficult to debug failed upgrades. It can also be extremely expensive, for example the leading Windows screenreader JAWS costs $900 per install.
And, like the article says, even leaving aside screenreader users this has a practical benefit for many users. If your prose is complex enough that you have to use a Latin abbreviation for precision you’re probably in need of a content designer to simplify it.
How does a screen reader translate FFS?
Abbreviations on the web can be expanded using the abbr tag:
<abbr title="Fat Finger Syndrome">FFS</abbr>
This also helps reduce ambiguity about the meaning of an acronym, improving the communication of semantic intent.
That said, it appears that expansion of <abbr> isn't consistently supported[1].
I suspect most screenreaders would simply say "Eff Eff Ess".
[1] http://www.powermapper.com/tests/screen-readers/labelling/ac...
> I suspect most screenreaders would simply say "Eff Eff Ess".
Correct, thankfully. I don't really see a use case for screen reader-specific usage of the abr tag. If everybody else has to read "FFS", possibly having to look it up on Google to know what on earth it is, why wouldn't I also? Although I suspect that as a sighted user, you can mouse-over some text marked up with abr and see the expanded form.
this doesn't address the complaint that non-native english speakers may not know what `e.g.` and `etc.` mean.
Presumably native latin speakers will feel right at home. :-)
I think they can't think of all situations (like people writing eg instead of e.g. or eff eff ess)
This would be my concern -- potentially half-assed engineered products significantly influencing nationwide (worldwide?) standards that suddenly many developers must now accompany. It'll be like supporting IE 7-9 all over again.
On GOV.UK we have high six figures of each of IE8 and IE9 per month, so we do support them already. It’s about user needs.
Supporting simple screen readers is a good thing, because it precludes much of the user-hostile crap that web ‘designers’ spew.
I don't think this will cost that much...
Is it actually officially correct to use "eg" instead of "e.g.", and "ie" instead of "i.e." in English? (UK, US, AU, or whichever else?) Both http://dictionary.cambridge.org and http://dictionary.com seem to show only "e.g." and "i.e." for me?
The Guardian style guide is the only one of the newspapers for which I could find an entry on this.
https://www.theguardian.com/guardian-observer-style-guide-e
It recommends dropping the dots. I know we were taught in school to treat them as abbreviations and therefore include the dots so I'm not sure where this new fashion comes from.
Maybe that's just the Grauniad codifying their well-known propensity for typographical error, though.
Just the current trend I think, hyphenation of words seems to have gone out of fashion too.
No, the correct form is "e.g." and "i.e.", and screen readers cope with those fine. People also get "etc." wrong a lot, writing it as "ect", which presumably screen readers would fail at, too.
Since people get it wrong, and there are more natural ways of writing these, these guidelines seem quite reasonable.
> People also get "etc." wrong a lot, writing it as "ect", which presumably screen readers would fail at, too.
You seem to be assuming that a screen reader's job is to somehow communicate meaning. It isn't. It's my job to listen and decide what I'm hearing. If you read "ect.", your brain automatically corrects it to "etc." because you're used to seeing it often. It's probably subconscious, right? It works exactly the same for me. I'm well-versed in the spelling/grammatical/other errors many people make when writing, the only difference is that I consume them via my ears, not my eyes.
Or, to put it another way, think about a guide dog. Its job is to actually guide a blind person, make decisions, keep them safe. It actually has to think. I prefer my assistive technology as dumb as possible though, so I use a cane. Its my job to use the cane, take in what it's telling me about the ground and area immediately around me, and quickly make decisions based on that information. The screen reader is more like a cane than a guide dog, basically.
> If you read "ect.", your brain automatically corrects it to "etc." because you're used to seeing it often. It's probably subconscious, right?
Not at all. Seeing errors like "ect", use of the wrong words like the all-too-common mix-up of "then" and "than", and even the omission of dots in things like "eg" and "ie" is very jarring to me. It usually pulls me out of automatic reading in order to consciously work out what the author actually meant. Even when I've worked it out, I find it hard to substitute the real meaning for what's literally there. Not that I necessarily blame the author, as I understand that mistakes do happen, but it does noticably harm readability for me.
I do find your description of your preference on assistive technology very interesting, as it differs from the preferences expressed by my visually impaired friends. They get varyingly amused and annoyed by their screen readers not understanding idiomatic things and reading them too literally. I wonder if the difference is that you're the kind of technical person that reads Hacker News, whereas they're both arty types who don't really do computers any more than the average person does. I suspect that were I to ever need to use a screen reader, I'd prefer the dumb ones like you do.
Ah, this seems to make sense, thx! Pity that this extended rationale is not present in the original article. Actually, this seems to render the article's claim rather unfair in laying the blame on screen readers ("[screen readers] read ‘eg’ incorrectly"), as it appears it should be rather "we have many errors of 'eg' instead of 'e.g.' in our texts, and they're too hard to catch for our editors"
edit: personally, I don't have to use a screen reader, but I do read words "aloud" in my mind when reading, and I believe seeing "eg" or "ie" without dots would make me stumble mentally too...
It seems that a lot of style guides these days actually recommend "eg" and "ie" over the versions with dots in a misguided attempt at simplification. I guess with those guides out there, the damage is already done.
Personally I always try to re-write sentences that use "e.g.", "i.e." or "etc.", as I think they're symptomatic of lazy writing. There's usually a much better way to write what you're trying to say. Similarly with excessive use of parenthetical clauses.
They could have changed it to their original form: "et cetera", "ergo", "id est"
e.g. is not ergo, it's "exempli gratia" (for the sake of example)
My mistake, thanks!
> Anyone who didn’t grow up speaking English may not be familiar with them.
Even fewer people are familiar with Latin.
"e.g." does not mean "ergo".
Would using the HTML <abbr> element with an appropriate title attribute help here?
So no more PM, AM, AD, BC, DIY, ASAP, LOL, Dr, PhD, OBE, KBE either I guess.
previously submitted https://news.ycombinator.com/item?id=12157284
i.e., "We've decided it would be easier to make everyone who updates a page on the GOV.UK website change their language than to make everyone that makes screen reading software produce a product that reads screens correctly"
> i.e., "We've decided it would be easier to make everyone who updates a page on the GOV.UK website change their language than to make everyone that makes screen reading software produce a product that reads screens correctly"
See my previous comments in this thread to learn why this is necessary, and why it's unrealistic to expect screen readers to "just work™". It's a nice goal, but it won't happen.