ICT and Real Lives: Big Objectives and Little People

Speech presented by KEVIN CAREY, Vice Chairman, RNIB and Director, humanITy, Techshare 2001

Date: 27/11/2001
Venue: Birmingham, UK

While it is important to emphasise the moral and legal cases for accessibility, we also need to work on the market case and recognise that, hand in hand with standard setting, access to information should be grounded in real lives, in a concern for banking, retailing and broadcasting. The prospects are good and we need to focus on success as well as challenges.

I am going to divide this presentation into three parts: first, I am going to say a few things about the visual impairment sector and IT, and the moral, legal and market cases for accessibility; next I want to talk about some of the technical difficulties and solutions to them; finally, I will make some concluding remarks.

The VI Sector and IT

Traditionally the VI sector has shared its attitude to information technology with the IT sector as a whole. We are definitely nerd-led; we are defiantly producer 'push' rather than 'consumer pull' people. It doesn't matter that most of our clients access information through televisions, radios and telephones, we think they ought to be getting into PCs and exchanging e-mails about the size of their hard drives. We have successfully, if somewhat uncertainly, got our PCs to talk interminably but our phone menus are dumb. We can surf the web using our PCs but we can't tune contemporary radios because the systems set-ups don't talk to us. We've got a wonderfully fine standard for describing pictures on the Web but audio description on television is a battle we still have to win. Who are we? And, much more important, who are our clients?

As I have said, we are only reflecting the generality of errors in IT. In this context, for example, it won't come as any surprise that the whole industry was shocked by the growth of text messaging which is used more than anything else for flirting; can you imagine a software writer flirting? We have become imprisoned by the engineers rather than being liberated by our marketing departments. In the case of visual impairment the problem is made worse because of the tendency of high school educated, congenitally blind, males to be gadget freaks as well as nerds. Not only do they ask me how big my hard drive is, they also want to know how many cells I've got on my braille display and they get all sniffy when I say "40". The sector is full of tiny niche technologists all heavily into producer 'push' and all primarily focused on the very high end of the market, unquestioningly buying into the assumption that computers are serous and worthy whilst television is vulgar and frivolous. It doesn't make any difference that the Internet is full of porn and television still broadcasts thousands of hours of worthy programmes about all possible manner of subjects, we all still look down on the medium.

Now I recognise that there is a serious concern for IT accessibility by blind and visually impaired people in education and employment but this is, demographically, a fraction of the market. It is also, let's face it, much more of a statutory responsibility which may explain the sector's interest in it; there might be some money at the end of little Jimmy's Statement which says he needs this and that in order to access the National Curriculum, and Gordon Brown is continually making statements about how far the Government wants all disabled people to work and that must include working with information; and you're hardly likely to get a generous response from Tessa Jowell, our Minister for Culture, if you say you want a government grant so that your almost blind uncle Bert can get much more vivid access to Internet porn sites.

So, with the market and the law as key components in any accessibility analysis, where do we fit?

These two pieces actually come together rather well. We can all make a moral case for accessibility and we might even make a limited legal case. This will be particularly strong in education and employment, as I've said, but not so strong in the case of your uncle Bert. We also have to bear in mind that the DDA was a more or less last ditch effort. We may get a bit more legislation tidying up some loose ends and plugging some gaps but I see no prospect of this Government or any other embarking on a major new piece of legislation. Whereas the United States has a written constitution on which claims to the enjoyment of rights can be based, we do not. There will be some case law on the DDA and on the European Convention on Human Rights, but no new big breakthrough; which is why we need to stop moaning and put some real money into test cases.

If we think that our clients really are getting such a raw deal we have to argue it in court, not in the New Beacon. More than 100 years after their widespread establishment, we still don't have a legal ruling on the obligations of public libraries to provide accessible material; and we've never even thought of mounting a test case, even if all our clients pay local and sales taxes. Everybody who is blind is currently paying for a library service they don't get. Blind people are cross-subsidising sighted people and what do we do? We beg for money from individuals and businesses so that we can make good the deficiency. RNIB may charge for its Talking Book service but that doesn't even cover the whole running cost, let alone the capital expenditure.

Now I recognise that that is where we are now and it will take a long time to get things changed; but must we make the same mistake with every new medium that comes along? We subsidise braille, we subsidise tape, Wireless for the Blind subsidises radios; will we soon be subsidising digital access? Where will it end? Is that what the voluntary sector is for? Is that why people put coins into tins and abseil down skyscrapers? I doubt it. Of course, there is no harm in our acting as paid agents for the public sector and major corporations but even then we have to be tough and charge the full rate. Then again, we can only charge the full rate if we are sure that we are efficient organisations that can fully justify the rate we charge.

We also have to make a serious distinction between our standard, flat offer of service and the frills that some people want. We are, as a sector, too bound up with providing a wonderfully bespoke service for 5% of the sector instead of providing a basic, rough and ready service for 95% of the sector. Should RNIB, for example, as a sector leader, establish a basic file format for digital, accessible information and then, with the exception of library books, expect the end user to sort out embossing? Why not specialise in realising file formats and providing tools such as a basic braille Grade One translator, and let the downstream people do the rest? This would mean more of every medium for everyone, including, crucially, large print, but it would force the sector to have a very different policy on braille embossing. This is a good example of both doing what you do best and putting yourself into a win/win situation, more file creation means more of every format of accessibility.

We also have to sort out the kind of relationship we have with our client group. At the moment, it's almost entirely that, a client group. We have made some moves towards thinking of our potential beneficiaries as customers. In this information age might we not want also to think of them as an audience? We currently provide people with exactly the information they want through books and documents. Many RNIB Talking Book and NLB users only want a kind of book, thrillers, romances, so we provide what I would call a 'middle casting' service. But as we can't serve most of our potential markets this way, why don't we become broadcasters? Why not have an audience for what we have to say rather than clients or customers? There is plenty of evidence that people with poor sight don't want to classify themselves as blind or visually impaired but they may need our kind of services. Because of technological development we can reach these people and measure the reach.

This is not going to be easy because our existing beneficiaries have a very strong grip on what we do. We have all fallen into line with the idea of our consumers becoming the key players in governance but what kind of representation of the VI sector as a whole do we get? Yes, late middle aged, white, male, congenitally blind high school graduate braille readers like me. Look at the struggle there has been to get any money in the sector devoted to large print or to the needs of the very old. If that means moving money from my kind of people, no chance; and of course sighted officers and committee members are transfixed like rabbits in headlights when this happens. We have a skewed relationship where somebody, just because he is blind, is allowed to say more or less anything and go unchallenged. This is plain silly and it is seriously biasing the services which visually impaired people receive. It has happened over large print but we should not let it happen again over access to all kinds of information in the information age.

Our central problem is to work out what the law and public sector funding will do and then to work out what the market can do.

The key concept to get embedded into thinking and planning is that we should only do what the general market cannot take care of. For more than a century the voluntary sector has more or less claimed ownership of blindness and been a caretaker for visual impairment. What began as pioneering services became established; some became public sector, others were the staples of fund raising. What, to return yet again to Talking Books, would the RNIB do if the Department of Culture offered to finance it 100% and run it direct through the public library system? Bang would go a huge slab of voluntary income, much greater than the current loss. If this really happened, a number of services the Government would not want would have to be trimmed or even closed. So the market, seeing the tin can and supporting its staff in a mass abseil, wonders why we are making such a fuss about access by VIPs to information.

A fascinating snapshot of where we are is provided by two pieces of work on disability. The first, which came out over a year ago entitled An Inclusive Future by Ian Christie, then of Demos, showed that approximately 1/3 of the population would be classifiable as disabled at some point in its life, mostly near the end. It was therefore foolish, he concluded, for disabled people to see themselves as cut off in a small group from the rest of society; society needed to learn to think of itself as an entity which a huge minority of disabled people in it; it needed to be inclusive. Many people who would become disabled would not be severely so but, still, if you are planning a service and a third of the people are likely to need something special it can hardly be classified as niche. This analysis was greeted by the official disability movement with extreme hostility; as usual it was the budget hoggers to the fore.

A few weeks ago the RNIB released its latest portion of the 2000 Needs Survey entitled Access to written Information by Ian Bruce and Mark Baker. Here are some of my favourite findings:

11% of the general public has difficulty in seeing a number in time to stop a bus. 14% of the British public have problems reading TV captions. And a quarter of the population has problems reading food packaging. Talk about market reach! Talk about extending our relevance! Talk about campaign cohesion, getting allies, accumulating support and funding for more esoteric projects. No, let's not bother.

I was at a recent discussion of these findings at a key RNIB Committee and the only thing the Trustees were really interested in were the statistics on the demand for learning braille; so you know where a quarter of the general population can stick its illegible food packaging?

This is slowly changing, TESCO, the British Bankers' Association and others, driven by the Web accessibility Initiative the RNIB, NLB and others, are beginning to understand the moral, legal and market case for accessibility and there is even braille on my bleach bottle; but are we united on the approach, or are we equivocal? What, we have to ask ourselves, will happen to us all if we succeed, if every cereal packet is legible, if every item on a supermarket shelf has the kind of voice chip you find in talking greetings cards?

The answer, of course, is that there is always something else to do. There is, in our field, the whole domain of broadcasting which affects just about every visually impaired person but which has had a low priority. I will come back to this point in just a minute but I want to round off the institutional analysis by saying that in this world of the public sector and the market the voluntary sector is inevitably left with two roles, as back-stops and as pioneers. It isn't pleasant being a back-stop, doing what nobody else wants to do, particularly if we are left with the kind of service provision that doesn't easily attract funding, so we have to work out how we can combine this with pioneering, which is much more professionally rewarding and has a better chance of attracting some funding though, having said all that, the main sources of funding are usually interested in supporting second generation services. Forget all this stuff about the Government and, indeed, the voluntary sector, being all in favour of cutting edge projects; the public sector and charities both hate it because of the risk, the risk to public and to donated money. But, still, if we are not to be left simply as backstops after the cherry picking is over we have to tap into the funding and energy sources for pioneering work or we will ossify.

Some Technical Observations

I said at the beginning I wanted secondly to turn to some technical problems.

First of all, it is absolutely vital that we begin to look at information systems holistically. There are three parts: information design (data); transmissions (carriers) and access (receivers). Let me illustrate the problem by referring to the history and current state of digital television and its use by VIPs. Once upon a time we were all very serious about audio description and we actually got a line into an Act of Parliament specifying legal requirements on information publishers but the law didn't apply to the set top box manufacturers so there was a stalemate for a while; there is a simple illustration of the manufacture and the reception not tying up. In the case of wanting people with poor vision to be able to shout instructions at their telly, the current problem is the data transmission, the carrier; currently it can shout at you but you can't shout at it. You can choose which Wimbledon game you want but that's the extent of television interactivity at the moment. So if we want a genuinely interactive system, particularly if, as I recommend, the information and the tools are all server rather than client side, then the design of the carrier is fundamental. We have behaved differently with different media. Because they are primarily voice-out systems we have assumed the accessibility of radio and telephones though both are now inaccessible to an important degree. With television we have pushed for content modification but were slow to recognise the receiver problem and I am not sure we've got our heads round the carrier problem yet. In computing we've been good at understanding basic data creation and the modification of receivers but we've been very bad at thinking about convergence and the move towards multimedia.

So here's the second point. The really important thing to focus on is the way that information is created. From our point of view we have to look at six key sets of factors:

Stratification (layering and labelling)
Translation (Migration between formats)
Navigation (choice of strategies)
Conversation (interrogation and response)
Customisation (heuristics)
Integration (of data and tools).

These are not in order of importance but approximately in the order they occur. I will quickly deal with each of them:

1. Stratification (layering and labelling).

The easiest way to picture the proper way to make visual information is to think of the way music is recorded; each instrument is put on a separate track and the whole lot is mixed together; but the point is that it can be unmixed and then re-mixed; or you can take one track of the recording, one instrument, and fiddle about with it while all the rest is left intact. At the moment most visual information isn't made that way, it's made like celluloid film where the whole image is stored on one, indivisible frame. If you take a can of blue paint and a can of yellow paint and mix them together you get a double sized can of green paint but you can't separate them out again; that's how most current information is made. What we actually need is the layering and labelling of information with each element placed on a transparent frame so that al the frames can be stacked to produce the final picture which can then be 'unbound' and turned back into layers. This way, you could isolate an element and magnify it, or leave everything as is except the background colour of the whole stack. In addition to helping people with low vision, this manufacturing process would make it much easier and cheaper to update information, altering text, for example, while leaving a picture intact.

2. Translation (Migration between formats)

Has it ever worried you how little large print has emerged from the use of digital files for braille production? Or, conversely, the tiny amount of braille that has been produced since most print production was digitised? The clue here is manufacturing information that can be migrated. This means making it in a very simple way that is amenable to tools; or, to use the contemporary manifestation of that, XML with style sheets. We need, for example, to separate our text from the page making function so that a tool grabs a chunk of text and headings and renders the pages according to our requirements. In the case of large print this might mean attaching images to keywords in the text so that as the print is enlarged and words migrate from one page to another they carry images with them. The whole craft is in the page making technology. In this area we have naturally concentrated on actual files, on data, but the migration of metadata is equally important, allowing a person with low vision to access metadata efficiently on a screen whilst a blind person uses a screen reader; what is good for one may be very bad for the other unless we've designed the right migration tools.

3. Navigation (choice of strategies).

The same kind principles apply to navigation strategies. Some of us are good at alphabetical or numerical lists; some data needs to be arrayed chronologically; but a home page with more than 100 links like that of the BBC isn't much use, even in text only form, if you want to access it with a telephone or remote controller with nine choice keys and the "0" for "Other". I have spent a lot of time lately classifying things into nines so that they can be used by numeric keypads on televisions and telephones but this should be one strategy alongside alphanumeric, chronological or key word searching.

4. Conversation (interrogation and response)

We are far too in thrall to the old PC based systems of communication that we are ailing to adapt telephone technique to everyday text operations. There is no reason why we should not design an information system that asks questions instead of simply listing some choices; and there is no reason why we shouldn't design systems with relatively simple micro vocabularies which would allow us to shout one-word instructions. If you are filling out a tax form there is a relatively restricted vocabulary, just as there is if you re filling out a questionnaire on soccer. Most of us use very small general vocabularies in tandem with quite big micro vocabularies for our jobs, our hobbies, the particular thing we're doing at the time. The temptation is always to combine networks together but this isn't scaleable for voice-in where you want the smallest vocabulary you can get away with.

5. Customisation (heuristics)

A couple of years ago I undertook some work which examined the routines for customising workstations for VIPs in the public library system. This involved hanging a screen magnification package, a screen reader and a braille display onto a PC bundle. We then watched VIPs getting to grips with the kit by asking them which media they worked in; whether they wanted red on black or blue on purple; what font they liked, what size of print. We had to shift text into the left hand side for screen readers and for people with poor navigation but we had to blow the text up, right across the screen, for others. On average it took us 90 minutes for each individual customising session and before Microsoft NT there was no efficient way of saving the settings. Now that we can do this we must. People should be encouraged to collect their own customised settings, store them in a system, on a disc or smart card or, best of all, save them on a personal website so that they can be downloaded into any on-line kit.

There are, of course, two things to remember here. Like fully sighted people who use a television quite differently from a small screen mobile phone, the VIP doesn't have a rigid primary access medium. Secondly, we need to customise depending on the output device. The best way to do this, of course, is to integrate the customisation with tools and I will come to this in a moment.

Meanwhile, the best way to cut down on customisation time and tedium is to invest more in heuristics, the ability of systems to adjust to their individual users. This is already eminently possible for most of our everyday needs and the concept of favourites and bookmarks is positively Flintstone. There are quite robust intelligent agent packages now which are particularly useful for VIPs who find scanning huge amounts of information so tedious. Some important applications of heuristics are:

Electronic Programme Guide Assistance
Study planning
Route planning and travel.

At a much lower level we have been really lazy about teaching our clients to establish macros.

The important thing about heuristics, however, is that they really need to be server not client-side applications so that we can use them regardless of the output device. You don't want to download your key user profile off the web into a screen at a railway station, you simply want the system to go to where your characteristics are and adjust itself.

6. Integration (of data and tools).

Which leads, finally in this section, to the integration of data and tools. We are far too rigid in this area. If I am in a difficult document I want to run parsing and précis, thesaurus and customisation as part of the integral system; none of this getting stuck and sorting out how to call for help. It's just like when you try to fill out an on-line tax form and all the assistance, so called, is in a different place; why can't the form itself ask questions and provide tiny chunks of advice for each item as default; you can always switch off the advice if you're an expert. The simple rule here, as with all systems, is that default should be for learners on the basis that more qualified people can modify or switch off.

This leads above all to one major conclusion. The future of ICT for all, including our client group, lies in server side intelligence. Those of you who have scraped a living out of screen readers, screen magnification, braille translation software, clever bits of turnkey hardware, should re-think your strategy. For everyone, including visually impaired people, the fall in real terms in telecommunications and the growth of the use of communications protocols like Blue Tooth mean that we will all soon only have two pieces of ICT kit; the thing you walk with and the thing that stays at home. These will be customised and modular. We will no longer each own four televisions, two radios, a couple of mobile phones, CD players, desktops, laptops and PDAs; all of these functions will be in these two central processors; one as the digital spider in our house, the other as our travelling companion; they will drive cookers and screens, diaries and printers. For some time yet there will be a need for a braille display module, though this should be no longer nor thicker than a 12-inch plastic ruler. Most of the rest of what we think of as access technology now will soon be redundant.

On the information design front, we need to spend far more time thinking about the way we build information from scratch; and we need to remember that there is no point having wonderfully accessible data if you can't get at it because the metadata is non accessible; that is rather like having a cigarette without matches; even worse than not having the cigarette at all.

You can see from this presentation so far that it is rather difficult to split the legal from the technical, the aspiration from the realisation.

Conclusions

You can tell a person by the nouns they use. If you go to a summit meeting of politicians and civil servants you hear a lot about milestones, benchmarks, good practice, beacons and exemplars, audits, comparators and parameters. In my world you would hear a lot of talk about banking, shopping and broadcasting, access to financial data, store inventories, health information, bus timetables and pornography.

1) Universal Standard of Service

So far we have been taken in by the technology. Another way of looking at this problem is to start from the output and work backwards. This is where TESCO and the British Bankers have been so crucial as developments.

I want to access a full range of banking services. Currently I walk to my branch and do my business. The bank branch is closed down and I am told that I can bank on-line. I plead that I don't have a computer. I am told this is tough. I do have a computer but the data is inaccessible because the information can't be manipulated. At this point the temptation is to get into a terrible technological haggle about why it isn't accessible and how it could be. My answer to all of this is to forget the technology altogether and develop the requirements for a minimum universal standard of service. It would be nice if I could choose between, say, a branch to which I could walk, a personal phone banking service, an accessible on-line banking service or (and this is an area we know very little about) an intelligent hybrid system.

2) ISP’s & Hybrid’s

The large VI agencies might, for example, combine to provide a combined ISP, banking and utilities service that combines the flexibility and personal help of a call centre with the privacy of on-line access. We might, further, rent this service out to those major providers who have an obligation to provide a universal service but who would prefer to lease this out to an agency. Such a system would involve considerable investments in accessible technology but the point at issue would be the service, who is obliged to provide it and pay for it, not some esoteric argument about why the set top box can't decode audio description strands.

3) Agents and Intermediaries

The agent/intermediary model is also important in the field of democracy. As Government takes decisions ever more quickly it is naive to assume that we can go on with the quarterly committee which makes decisions, or more often, sends them down to a working group. One of those ironies of the new technology, like the inaccessible first generation of digital radios, is that it is actually diminishing participation for disadvantaged groups like our clients. The answer to this is to develop a new theory and practice of intermediary trust. This might mean the establishment of an official Register of Trusted Intermediaries and it would certainly mean that agencies of and for the blind should look very seriously at developing themselves as trusted intermediaries. This is why the RNIB Membership scheme is so important but it needs to be combined with the personal contact of the local societies.

4) Trust

This, however, requires yet another development. Blind people in general and the more activist sector of the VI in particular has to learn to stop mistrusting the organisations which exist for the sole person of improving the lot of the client, beneficiary or customer. We have all made mistakes in the past but if anybody is to blame because blind and visually impaired people are poor, unhealthy, unemployed or lonely it is the fault of the public sector, not charities; that's what we pay our taxes for. In the past half century the VI world has almost wrecked itself in a seriously damaging and stupid civil war while the Governments of various political stripes have been able to leave us to ourselves. There was a little post War assistance from Lloyd George and then Atlee; there was the heroic Alf Morris under Wilson and there was the DLA, so skilfully extracted in the Thatcher years and latterly the torso of a DDA; but it's precious little to show for fifty years of activism. The disability Stalinists have been too interested in ideological purity and keeping their own tiny bands of troops in order. We, as a sector, have to recognise that our future lies in working with an indifferent Government, working with an ignorant industrial sector, in order to get what we want. We need to learn, as industry has had to learn, how and where to collaborate and compete.

The need to do all these things is surely summarised in the title of this event. I like the name "Techshare", it stands for all the things I stand for and I hope it stands for all the things that the RNIB stands for. Sharing does not mean you lose your individuality; pooled sovereignty doesn't mean losing control. Sharing ideas doesn't mean that they are being robbed for exploitation by others. Defensive intellectual property strategies are bound to fail in an open world. You are only as good as your next idea.

Let us, in sharing, in providing our expertise on new technologies, always remember that our big objectives have to have an impact on the real lives of small people.

KEVIN CAREY
Vice Chairman, RNIB and Director, humanITy

28.xi.01.

Return to Accessibility General