Monday, July 26, 2010

The Questionable State of Editing

Yes, I think I have earned the right to critique editing. I have been producing writing on the computer to be edited for at least 20 years; and I have been editing others’ work for more than 15 years. In my youth, I thought I might be a writer; so, in script, by typewriter, and within word processors (one of which I helped design) I have been editing myself and seeing others’ comments on my writing for more than 50 years.

However, I don’t agree with the usual criticisms or defenses of editors that I hear. I don’t fully agree with the dependence on Strunk & White; I do feel it has improved writing. I don’t believe that writers are typically superior to editors in their judgments; I don’t feel that editors automatically improve written material. I don’t feel that the greater ease of writing and the automation of spell-grammar checking afforded by word processing software has decreased quality or “blandified” the results; I do feel that the ability to produce more writing in the same amount of time has increased the perception by the reader that some of it is repetitive.
Let me summarize what I do believe:

1. Editors frequently fail to see that the editor’s job is both to make an argument or narrative clear and attractive to the reader and to preserve the distinctive voice of the writer.
2. Both writers and editors typically do not fully appreciate the importance of the general layout of the written material on a given page.
3. Editors tend to wrongly value completeness above conciseness, because they fail to consider the effect of cumulative particular changes on the whole piece.

The Editor’s Job
Here’s a story from my days at Aberdeen Group: I proposed to start a piece on IBM’s AS/400 product with the sentence: “Integration is In; and the IBM AS/400 has it.” IBM marketing people loved it; but the Aberdeen editor refused to allow it. He asserted that I would have a preposition with no object – clearly a violation of style guides – and that it was incorrect to say that the AS/400 ‘had integration’ – rather, it had features that enabled integration of customer applications – and that the semicolon should be replaced by a comma – clearly not by an em-dash! When informed that the newspapers in those days were full of lists of people and products that were In (meaning in fashion) or Out, he noted that European readers would not understand. In the end, as I remember, we ‘compromised’ with the sentence: “Integration is in fashion, and the IBM AS/400 has the features to enable effective use of integration by its customers.”

Now, the point of this story is not to pick on this particular editor, who was doing his best to ensure a common tone and higher minimum writing quality for Aberdeen publications. Moreover, I am not saying that the common sin of all editors is to squelch the individuality of the writer. On the contrary, I have seen many editors who, aside from enforcing the style guide, leave the use of passive voice or lengthy paragraphs pretty much intact.

My point here is that it is important that editors do both: make the piece communicate clearly to the reader and give the reader a sense that one distinctive person – and that person is the writer, not the editor – is talking to them. The editor’s criticisms were, by and large, correct – but I am pretty sure that the piece would have been better as is.

I think that what underlies this problem is a “consensus” view of readers – or, as Dale Carnegie used to say, the one negative thing we hear tends to outweigh all the positive things. It is true that some readers would have noted the lack of a preposition; but most would have enjoyed the In-joke. It is true that European readers would not have understood the allusion; but that would have made them more interested in finding out what the product was. It is true that the sentence was not clear as to how “integration” affected customers; but those that did not know already would find the sentence arresting enough to make them more curious about finding out – and the rest of the piece, not to mention the rest of the first paragraph, would have been ample explanation. Still, I think that the objections of the few outweighed in the editor’s mind the greater positive reaction of the many.
Now consider the same editing process applied to a whole piece. Most colloquialisms are replaced by words understandable by the densest reader; indications of speech patterns are replaced by rigid subject-verb-object constructions with few subordinate or follow-on clauses; each sentence is clear from immediate context. The result is a piece that is as much in the editor’s voice as in the writer’s and therefore is not as distinctive to the reader as either. In addition, the impact of the piece – its punch, if you will – is less; unnecessarily so.

By contrast, imagine the other editing style (“letting the thousand flowers bloom” by doing minimal damage to the writer’s voice) applied to a whole piece. Yes, portions of the piece are arresting; but the reader simply cannot follow what the writer is saying, often because the writer him/herself has not really thought out his/her argument. Letting passive voice go in all cases because it is “the usual practice in technical papers” or because it sounds more colloquial; treating ellipses as sacred when they are often slipshod memories of catch phrases; passing over long sections and long paragraphs completely – the result of these editorial tactics is often happiness on the part of the writer, but not on the part of the reader.

It may seem that I am asking the editor to walk a narrow path. No. I am asking editors to put themselves in the place of the writer; to ask, what is the writer attempting to accomplish here, and then to make the minimal changes to make that happen, while attempting to make those changes conform to the distinctive style of the writer, not to a one-size-fits-all template.

The Importance of the Visual
Once, when I was at Prime Computer, I was writing sections of a software design document that really needed architectural diagrams. In those days, the editing software available to employees did not have a PowerPoint or Visio to create such things, so the alternative was what’s called “line art.” You would use the characters on the keyboard to approximate the lines, circles, and arrows of a diagram; and because there was no kerning, the results were pretty viewable.
I got so good at line art that I started using it once per page of text. When I did, I discovered that “mini-graphics” improved all manner of writing. The visual was much more concrete to the reader; and because it was attached to almost a page’s worth of writing, the text explained the writing and vice versa – the two were in balance. Above all, visually, the graphics nicely broke up page after page of dense technical writing. Effectively, the document was more readable.

At an earlier time, I worked for a company creating the only word processing software I know of that typically displayed a page at a time. The result, for the writer, was to make him or her very conscious of the boredom the reader felt on seeing a page that was just one long paragraph. Readability improved, because the writer took care to insert bullets, numbered lists, and one-sentence paragraphs to “jazz up” the visual appearance of the text.

Since then, I have found relatively few editors that really appreciated the importance of the visual appearance of a page. Yes, most now require that a paragraph have no more than four sentences; but then they typically forbid one-sentence paragraphs. Many dislike bullets on sight; others demand that graphics be large, so that the text of a page is mostly crowded out. Tables are strictly rationed. The result, after editing, is page after page of 3-4-sentence, complex paragraphs, interspersed with a few pages of graphics that either are too vacuous and “marketing-lite,” or else visualize the data, not the key idea. The whole is then dressed up with “cut-outs”, to reinforce the idea that the text is not worth reading. And this is in writing that is intended to market a product or an idea; where that isn’t so, in histories or novels, the graphics are removed or isolated, bullets are verboten, tables are déclassé, and explanatory verbiage mushrooms to fill the gaps – making the problem even worse.

Now, granted, some of the problem is that today’s software just is not very good at supporting mini-graphic creation or odd page layouts – although it’s pretty good at bullets and tables. But I believe that the main problem is that writers know that editors will typically not accept their experiments; so they don’t bother.
To my mind, the solution is actually pretty simple: the editor should imagine a particular page, not with words, but with shadowed blocks where the words and graphics are. What looks more visually interesting, 3 large blocks of text on a page or 2 large blocks and 3 indented lines for a numbered list? What looks better, 2 large blocks of figures plus a thin strip of text at the bottom, or 3 larger blocks of text with a small block for a graphic interspersed? I know what I think; and the reaction I get from readers who are not conditioned to expect a certain layout is almost invariably positive: better visual layout, well balanced with the text, adds punch and clarity, no matter what the content.

The Forest and the Trees
When I was writing for Aberdeen Group, I had to write literally hundreds of white papers 8-12 pages in length that vendors would consider buying. As a result, literally hundreds of times, two or more reviewers on the vendor’s side would edit my drafts. And as a result, I learned to write short, choppy sentences everywhere in the first draft.

Why did I do that? Because the natural tendency of the editors was to perceive what was left out, and to request additions. And since the editor had strong control over whether the piece was published or bought, there was strong pressure on me, and on every writer, to accommodate additions, while still somehow preserving the overall flow of the argument and keeping the visual layout and organization simple. That could only be done by cramming more and more clauses within each sentence. By anticipating that editors were going to do this, I was able to avoid massive reorganizations of the white paper that would have increased the likelihood of white paper rejection. However, the results of several rounds of editing were not usually very pretty.

Here’s an example. The initial version of one paragraph about Oracle OLAP Option was:

Within the general class of analytic tools, OLAP differentiates itself by its ability to capture data in sophisticated “multidimensional” ways, as compared to relational data, and to perform “what-if” analyses on that data. Therefore, OLAP is typically used less for standard reporting and straightforward querying tasks, and more for sophisticated analysis.

After the thirtieth (!) editing pass, it became:

Businesses are increasingly finding that traditional query and reporting tools are not enough. Traditional query and reporting tools are sometimes not performant or scalable enough to support all of users’ complex ad-hoc queries, not efficient for analysis beyond two or three “dimensions”, and not functional enough for sophisticated trend analysis or forecasting. For example, a CEO of a Value-Added Reseller (VAR) might have observed a significant downturn in business sales during the most recent month. Using a query tool, the CEO would typically look at revenues for that month for all regions, possibly via a fairly complicated query. Such a query could take quite a while to return results from a very large data set. By contrast, a good OLAP solution enables rapid, in-depth analysis and forecasting involving difficult-to-anticipate ad-hoc queries for both immediate tactical and longer-term strategic decision-making. OLAP also differentiates itself by its ability to analyze data across more than two or three dimensions.

Actually, I’ve left out the fact that I had to subdivide the above into 8 sentences, 3 paragraphs, and 2 sections. And chop out later material to make room. And reorganize all the sections to accommodate the additional section. And that was only one paragraph …

To an editor, it looks like the piece is becoming better. There’s more information in there. There’s more reinforcement via examples. Because the argument is laid out more fully, it seems more credible. There’s just one problem: because of the added length, the most important point you’re making is less clear. By planting new trees, you have obscured the overall view of the forest.

This relates to my view of the effects of new writing technology. Compared to typewriters and writing by hand, I love word processing. No more desperate avoidance of typos; no more carriage returns; no more slow recording of thoughts; no more rigid organization determined by the first draft. And the fact that you can turn out acres more text is fine by me: I find that what I lose in careful scrutiny of each word I gain in the ability to instantly see and correct for gaps in the argument or the narrative. No, I don’t think that the average product of the word processor is worse.

However, the editing can make it worse, or reinforce its bad tendencies. Handwritten and even typewritten works from the old days are denser, more concise, because every word counts. The experimental, the elegant, stands out more clearly. By contrast, lazy word-processor writing can seem much more the same, with the same old stylistic quirks repeated ad nauseam. And the tendency that I have cited above, the tendency of editors to value “completeness” over “conciseness”, makes those features of the new word-processor-driven writing worse.

To me, the solution is for editors to recognize, deep in their bones, that there is a tradeoff here. The way to see if you have gone too far is to go to the very first paragraph of the piece, where “why you the reader should care” should be laid out very clearly. Then go to the offending section, and ask, is it clear how the edited version fits with the argument laid out in the first paragraph? Can I read the first sentences of each paragraph in this section and see an overall argument that makes one key point within that first paragraph? If not, is the added information really worth the cost in clarity?

In case you haven’t noticed, in this piece I have been practicing what I preach (as I feel appropriate). There’s colloquial speech in this post, plus paragraphs mostly limited to 2-4 sentences. There’s a bulleted list on the very first page; there’s variation, and experimentation; I have limited my “completeness” in the interests of clarity. It seems to work for me.

But if you’re not convinced, here’s another example. The first paragraph of a draft I recently submitted for editing runs as follows:

Roger Zelazny, in his novel “Lord of Light”, had one of his characters comment: “None sing hymns to air; but oh, to be without it!” Run-the-business applications like xxx are an organization’s air supply: without effective performance management, response-time delays result in angry customers and unproductive employees, cause outages from trial-and-error fixes, and signal problems that will lead to outages – making xxx effectively unavailable for long periods, choking off the sales and production cycle, and causing the organization to seize up as if it were gasping for breath.

Now here’s the same thing, after editing (that I had to clean up a bit):

The use of BI (business intelligence) software for speeding the organization’s financial reporting and planning/budgeting has an unusually direct impact on an enterprise’s bottom line. Closing the books rapidly means that the CEO can immediately react to changes in the business environment, and avoids money wasted on failing investments. Faster planning, and budgeting according to a “what-if” plan, allows deeper investigation of alternative strategies, ensuring that the final strategy will be more effective. Planning that can take into account monthly and quarterly results and implement a new spending plan within a week of a period’s end is the mark of an agile, cost-effective organization. However, complex organization charts for growing organizations doing deeper analysis mean that BI planning/budgeting/consolidation software must perform and scale better with each version.

Which makes the reader care more about reading the rest of the piece? Which makes the main point of the piece – you need good performance management software – clearer? Which grabs the eye even before you’ve really started reading?

You be the judge. 