Revamping writing skills is a major phase in vision loss transition
One reason for starting this blog was to regain my writing skills. This post describes my personal techniques for writing while using a screen reader and other assistive tools. A suite of recorded mp3 files illustrate some steps in rewriting and expanding the previous post on Identity Cane.
Most of this post assumes a state of experience comparable to mine three years ago before I became print-disabled. It was hard then to know what questions to ask to prepare myself. I bumbled through using the TextAloud reading application which enabled me to write well enough while I could control the lighting around my PC and begin to experiment with alternative screen reader packages. Unfortunately, I had some truly humbling experiences trying to edit rapidly at review panel meetings with overhead lights bearing down, voices all around, and a formidable web-based panel review system. Following the edict "Do no harm" I recognized a challenge of physical, cognitive, and technological dimensions. I had to admit I was professionally incompetent when it came to writing, ouch!
My model for writing without vision
The basic questions are:
- What are my accuracy versus speed trade-offs? And, how do I manage them?
- What tools do I need? And, how do I teach them to myself?
- How must I change my writing style? What are the new rules of ‘writing by ear’
If you are not sure how this writing process is working, listen to me writing some text using the NVDA screen reader.
The tradeoffs of accuracy and speed
The Accuracy Versus Speed Tradeoff is intrinsic to writing. How fast do you record your thoughts, accepting some level of typing and expression errors, with separate clean-up edits and rewrites? If I type very fast, I make more errors but am better able to record the thoughts and even establish a "flow" mental state. Writing more slowly allows corrections of wording, punctuation, and spelling but risks loss of thread and discouragement from a feeling of slowed progress.
Writing and editing are very different cognitive tasks complicated by operating primarily in listening mode. The input and output parts of the brain must operate together. A document filled with typos is pure agony to correct, causing a cascade of further errors and often destroying the structure of the whole document. One twitch in a edit can remove more than a letter, even a line, sentence, or paragraph. In "computational thinking" terms, the trade-off is to design the interactions of two concurrent processes that interleave events and actions to produce a document with an optimal amount of errors to be removed by even more processing involving editing tools.
I tried several drafting techniques. Writing in long hand notes, outlines, and snippets had worked for 40 years but I could no longer read my hand-writing. Recording into my Icon PDA helped organize my thoughts and extract some pithy phrases from my brain. As my memory has improved to take over former vision-intensive tasks, I have found it possible to mentally compose a paragraph at a time then hold it together long enough to type into the word processor.
Basic writing and Listening Tools
Writing without looking requires several tools, with my choices discussed below:
- Compositional, for typing, formatting as needed, and editing
- Spell checker, possibly a style or grammar checker
- Pre viewer to present the written results as they will be read by sighted, partially sighted, and blind readers
- Speech tools to read while typing and editing, as well as presentation of the written result
- Voices to capture alternative audio presentations of written results, as well as feedback on style and tone
My personal process is:
- Compose in mostly text with minimal HTML markup using Windows NotePad;
- Use the NVDA screen reader for key and word echo, with punctuation announcement off then on;
- Copy text into the K1000 tool, applying its fabulous spell checker, listen for errors and speaking flaws using its self-voicing reader, and copy back to Notepad;
- Listen in several voices, including both female and male, for flaws and nuance of style;
- Preview in a browser, Mozilla Firefox,to grasp whatever I can see on a large screen and to check links;
- Copy into wordpress blog editor.
The obviously best choice for writing is the word processor most
familiar to the writer. However, criteria may change as vision degrades. The spell checker may not have visible choices and may not announce its fields to a screen reader. Excess interface elements and functionality can get in the way. Upgrades and transition to a new computer may demand new software purchases. After years of Microsoft Word and Netscape HTML Composer, I finally settled on the combination of Windows Notepad and Kurzweil 1000. The trickiest feature of the ubiquitous Notepad is "word wrap" for lines with very few other ways for a writer to screw up a document. Since I write HTML for my website and blog, using Notepad avoid temptations of fancy pages by not using WYSIWYG. Also Notepad never nags for licenses discount deals, and upgrades,
On the upscale side, I needed a scanner manager for books and Other printed stuff. The Kurzweil education Systems 1000 offers not only scanner wrappers but also several word processor features. One is a beautiful spell checker to read context, spell the word,offer alternatives all using its own self-voiced interface. Listen to me and the K1000 spell checker. I also like having a reader with alternative word pronunciation, pausing, and punctuating. However, I occasionally lost text due to lock-up and unpredictable file operations, so I opted for the universal, simple Notepad for composition.
Update December 2008. I am now using the free Jarte editor based on Wordpad. Behaving like the Windows Wordpad, Jarte has a spell checker similar to K1000, multi-document management, and other features. Most importantly, the interface recognizes and cooperates with a screen reader, NVDA for me. Carolina software designers have done a great service for visually impaired writers and should serve as a model for interface developers of other software products. I’ll be upgrading soon to pay for the free version and some extra features.
A screen reader drives writing. by listening
As discussed in NVDA screen reader choice posting, I do not use the conventional expensive screen readers in favor of a free, open source wonder the I expect to rule the future of assistive technology. NVDA allows me to switch among voices, choose key and word echo, and degree of punctuation announced.
Writing and reading by listening has surprising consequences. First, it strongly differentiates sighted readers from those listening who will probably not hear the colon you use to start a list of clauses separated by semi-colons. Second documents must be read multiple times, with and also without punctuation announcements. It is difficult to concentrate on the sentences when every comma, quotation, and dash is read. And it is necessary to hear every apostrophe and other punctuation to locate extraneous as well as missing items.
Synthetic voices alter writing practices
Another suite of editing tools are synthetic voices, which may come as a surprise to many sighted as well as newly unsighted writers and readers. Synthetic voices have dictionaries of pronunciations but inevitably screw up in certain contexts. Is that "Dr." a street or an educational degree title? Is "St. Louis" the city with a saint or a street? is 2 the numeral like two spelled out or too as in also? No matter your screen reader settings and data, your readers may differ. Well, some of this can be tweaked but generally my attitude has been to just live with quirks.
Synthetic voices offer an even more powerful editing feature unknown to most sighted writers. The excellent researcher Clifford Nass" "Wired for Speech" tell how our brains react differently to gender, ethnic, age, personality, and other features of synthetic voices. Even if we know the voice is only a data file, we still confer more authority to male voices and react negatively to perceived aggressive female voices. This allows editors with synthetic voices to identify phrases with a tone that might be perceived as weak, over-bearing, age-related, or introverted. Don’t believe me? Listen to examples of male and female voices.
Note to sighted writers: you might also find these techniques assistive for finding typos, checking style, and evaluating the forcefulness of your writing. Nothing says you have to be visually impaired to try writing by listening.
Complexity becomes more visible with vision loss
When I write my blog, I must address both sighted and unsighted readers. Sighted people see a dull page of text, while people listening to the page or using magnifiers or contrast themes may react differently to a posting on a myriad of textual, graphical, and audible facets. Much of this out of my control as I cannot see the appearance of my pages in your browser, nor do I know if you are listening in a browser or an RSS client. Also, your speech settings, if any, may differ from mine in speed, dictionary, gender and more. .
A very insightful article on writing for accessibility points out the ill effects of complex sentence structures, reliance on punctuation, expectations of emphasis, and unawareness of the span of settings possible on the end users side.
Now, in my technical and business writing days, I was the "queen of convoluted sentences". I just never understood what was wrong with sub-sentences (as long as the sentence parsed ok); rather, I thought it a mark of quality. Whoops, there I did it again. I used a parenthetical phrase that might not be read with parentheses around it. And I relied on a semi-colon to separate sentences. Sorry about that, I’m working hard on this. But, there I made another mistake. I used a contraction which synthetic voices have trouble pronouncing “I’m” when I could say "I am". Abbreviations are also problematic. Should I say ER or E.R. or "Emergency Room’? This is giving me a headache.
The strongest lesson about compensating for vision loss is that ‘Complexity really hurts’. Overly complex things, whether physical or informational, cause accidents and invoke recovery methods. All this wastes precious physical energy. It is easy to be discourage when tasks that could be performed before vision loss are now too expensive in energy or time. But, conversely, I can now see complexity for what is, usually bad design. And, on the brighter side, once the source of complexity is identified, there may be a work-around, a simplification, or a suggestion for a better design. All this conscious adjustment of expression practices may actually be good training for aging more gracefully. Sigh.
Recordings to Illustrate Writing by Listening
The following recordings accompany this posting. Mp3 files may download or launch a player, depending on your browser and computer settings.
Listen to me writing shows the screen reader speaking text in Notepad as written and revised.
Spell checking and listening in K1000
Listening in several synthetic voices for gender and other differences
- Audio version of this and other posts