Before I start to lose more readers than gain new followers let me give a very brief, and far from incomplete but still conclusive overview on some technology that helps science education. I stated earlier it is important to notice the significance of substituting sight but it is not enough to only recognise the problem. We need to act, rather than just discuss and say heroic phrases.
The baseline of the challenges behind doing mathematical sciences, – or in actual fact any science that involves any form of mathematics such as physics, chemistry, astronomy but even psychology, economics, engineering and so on – is the scientific language. Language is the way we create, record, share and inherit knowledge throughout generations. Regardless the style of communication, the modality of conveying a piece of information, we do it by means of a system of symbols. Depending on the local culture of a geographic region humanity uses various signs to symbolise given thoughts. We can draw these symbols, we can verbalise these character strings, we can sing them, convert into body movements, a set of light fleshes and so on. Mathematics has its universal language across most cultures on Earth, but what it also has, a unique set of symbols to represent numerical information and the operations on those values.
Taking the most conventional and popular way of creating and distributing scientific knowledge i.e. writing and reading, we face the first problem. How do non-sighted people read and write? There are a number of ways, two of which are the use of touch or hearing. You can transcribe information into tactile formats, not necessarily braille dots but simple embossed shapes, or you can speak the information out loud. You can do it, I can do it, people near by can do it, we all are able to communicate applying touch or hearing. Now the second problem is to automate the process of converting visual text into tactile or spoken format. To maintain individuality, flexibility and time efficiency this is done using tireless machines. This immediately brings us to the third and original challenge. Machines don’t learn easily and have difficulties seeing what is written if it is something more complicated than text composed of Latin characters.
Mathematical language has two key features causing the trouble in automating the visual to tactile or auditory conversion.
- It is full of special symbols implying an operation of some sort,
- It is multidimensional.
What I mean by the latter characteristic is the variation on the notation of subscripts, superscripts, writing numbers on the top of each other not to mention objects like matrices, vectors, or the use of the Greek alphabet.
It is no easy task for a machine, nor for the developer and designer of that device to deal with the number of different symbols to be recognised, or how to render properly the various positional notation from a visual input to a verbal, auditory or tactile format.
And then we have only covered textual information so far ignoring graphical information which became inevitable in science since Descartes. Modern technology makes it possible to read and write standard text even if someone doesn’t see. There is also great progress in rendering the mathematical content and displaying it very interactively using synthetic speech or tactile surfaces. One of the remarkable efforts made is the standards introduced as MathML, which is a “linear” way of writing and communicating information on the internet, such that it offers a visually appealing output for sighted users, but also enables text-to-speech engines to process the language correctly. MathML is becoming more and more available and also it is supported by the most recent versions of screen readers such as Voice Over on Apple products, JAWS for Windows from Freedom Scientific or other programs with web browser plug-ins e.g. Design Science‘s software MathPlayer. The other remarkable project – that serves offline source support more effectively when it comes to reading and writing mathematical documents, is an initiative I am also involved in, called LaTeX Access.
Further more, for those of you who might be familiar with JAWS (Job Access With Speech) surely the virtual keyboard rings the bell. It helps inserting special characters not present on the physical keyboard into text. The idea is great and it is done but guess how many symbols you have in the list of choice! The answer is 17 in one particular version and if I remember correctly from the top of my head, 33 in an other version. Even this little selection appears to be composed of utterly useless symbols. After a years search I finally found a way how this list can be extended. So the plan is to add a few hundred more characters to it, including mathematical symbols and letters of foreign languages. I want to do what should have been done a long time ago by professional developers. The technology is there, we know how to do it, just a bit of time and effort is required to bring the most out of what we have so far. Of course when inserting those characters as an example, we have to make sure they can be also interpreted in a non-visual format. This was not the case with the Greek alphabet until two years ago. Figuring out how easy it would be to make the TTS engine announce alpha, beta, gamma and their friends, I worked it through, coded it and surprise-surprise my own computer learnt how to speak Greek letters in less than a day, making mathematical text much more understandable.
This is just one of the many similar simple examples. What I really want to emphasise is that very often all we need is some ingenuity, creativity and effort in using and further developing the technology we are already served with. Thus I am less concerned about making mathematical language accessible. Using MathML, LaTeX Access, Voice Over, JAWS, NV Access, MathPlayer, Infty Reader and a handful of more digital tools, we can make reading and writing of mathematical content more convenient for people who might just have lost their sight regardless of how well they are equipped with coding skills. I believe in the near future it will be a fairly resolved challenge. The technology required is there, facilities are in good development for rendering text and equations with a variety of notations, and also standards are spreading ensuring the existence of a unified system. Although we are at the beginning of a journey,
publishers of science books pay more and more attention to accessibility and great initiatives such as Bookshare become a reality. However, what concerns me much more is the access to scientific graphics. This is a topic for an other day but certainly worth mentioning.