Over the past ten to fifteen years, countless people have been displaced from their posts by the ubiquity of computers; there aren't many office tasks that can't be sped up, and therefore made more "efficient", by the little beige boxes. You won't be surprised to learn that proofreading is no exception, but the principal danger does not come from the most obvious of sources - spellcheckers - but from other aspects of the publishing process. Before I go into the details, we'll take a brief look at the traditional, pre-electronic, process:

1 the draft is handwritten and edited;
2 this is converted into legible type by a typist;
3 the typescript is edited by a copy editor;
4 the corrected copy is given to a typesetter to set;
5 sample proofs are made;
6 proofs read by someone (job title forgotten);
7 corrected proofs given back to typesetter to reset;
8 galleys given to publisher/author who check once more;
9 final, minor corrections/revisions made;
10 the work is published.

Now, when most writers have their own desktop publishing facilities, virtually all of the above steps can be completed from the office, making a process as shown below:

1 the draft is written and typed by the author;*
2 the floppy disk and a printout are checked by an editor;
3 the typesetter puts the contents of the disk into the computer;
4 the editor's corrections are assimilated into the text and galleys made;
5 galleys returned to publisher;
6 final checks and revisions carried out;
7the work is published.
*Just because the author might know a lot about nuclear physics, it does not necessarily follow that s/he has a good keyboard manner or a perfect knowledge of the written language, which might not be his/her native one; editing is essential.

As you can see, the lowly proofreader has been squeezed out somewhere along the line (somewhere between stages 3 and 4 to be exact); the work does not need to be proofread as there are no proofs! The whole work is in a data file; the typesetter cannot make any mistakes short of missing out some of the editor's instructions, because he has the original there in front of him. One of the failings of the traditional system is the "Chinese Whispers" nature of it: an error made early on could be reproduced (and new ones incorporated) as the process goes on. There is little doubt the new way produces a more accurate result.

So that's the end of the argument? Well, not quite. The vast majority of copy editors have served "apprenticeships" as proofreaders, probably for several years; it gives the mind an editing bent to augment comfort with the language. There is more to proofreading and copy editing than knowing English, however; it's surprising how little one does appreciate about the way literature is put together until one has seen it in its embryonic stages - there are so many things that the book reader takes for granted. Courses are fine, but being time served adds a further dimension. It is a problem that I foresee growing as digital processing becomes ubiquitous, but that's not to say that the new methods are worse than the old.

There can't be anyone in the correction game who hasn't had this typo conversation at least once a week for as long as their word-addled, pun-lovin' minds can remember. I find a polite re-education more instructive than an act of violence. Others find apolite re-education more appropriate. Run that one through your spellchecker! The a- prefix can make any word into its opposite thereby giving the whole sentence its opposite meaning; now and not are two words often mis-keyed, which also can invert the meaning of a sentence (try it). Sentences like this one don't make good sense can slip through the net. Words with similar spellings butt different meanings will survive too, and even though my grammar checker doesn't seem to know the difference between that and which, it seems obliged to give me its opinion. All they are useful for is finding words such as flubblit that patently don't exist, which is why I always use mine, but never rely on it.

Everything these days happens quickly. Information is so easily put into the public domain that it can be published or broadcast without a second thought - after all, tomorrow will it might be out of date, and the message is more important than the grammar. The Guardian used to have a reputation for being riddled with spelling mistakes but its journalistic integrity was rarely challenged. But today's news wraps tomorrow's fish and chips, as the saying goes. With the Internet, anyone can put their opinions or their writing up for the world to see in seconds; as long as the reader understands it, who cares if there are a few errors? Not I. Does a person's spelling ability run parallel with her intelligence? No; indeed, the opposite is often the case. Which is more important to a surgeon; an education in English or one in biology?

However, it's hard to imagine companies trying to look professional when their websites or advertising literature contain basic errors, as the implication is that if they don't care about their own image, they probably won't care about their products or services. This may or may not be the case: does the grocer who sells potato's sell better vegetables than the one who sells potatoes? Does it matter or not? The truth is that it's subjective: the people who do know whether or not it matters are the customers.

Once work has been produced, it can be reset with revisions simply by scanning the original and having the computer read the shapes and convert them into new type, rather than a bitmap image. This way, the only parts that need checking are the revised ones. Sounds perfect! However, one or two problems can occur. Firstly, the accuracy of the output depends upon the quality of the original - both grammatical and physical. Cramped text can make, for example, "d"  be read as "cl"; sometimes "0" might come out as "o" or "O"; semicolons might be read as "i"s and problems with bold and italic type can be encountered. Secondly, the piece will still need formatting as exact levels of headings, changes in font and tracking, and various setting issues will be confusing to the computer.

Imagine a world where your doctor scrawls your prescription on a piece of paper which is fed into a reading machine which automatically dispenses your medicine. Would you take it?

You know, some of us would like it if technology never moved on - we could listen to our vinyl LPs, drive our model T Fords, and have monks write out our Barbara Cartland novels by hand. Some people see progress as a hindrance, and, in the case of vinyl, not progress at all, and these people are entitled to their opinions. Eventually we have to down weapons and live together, bounce off of one another, and make the world a better place. Sometimes, though, promises of technological advance are empty and unfounded, and the older methods really do turn out to be better. Some labour-saving measures cut the effort of the job itself but shunt more work on to another party. In the end it's not my place to make judgements. Besides, I use Macrosoft Judgement Assistant  version 2.0.